i9606

Community Intelligence session at ISMB 2012

2012-01-24T08:16:00.000-08:00

Looking for something to do in Long Beach, California on the afternoon of Tuesday, July 17, 2012 ? Stop by the special session on Community Intelligence in Bioinformatics at this year's conference on Intelligent Systems in Molecular Biology (ISMB)!

We have four distinguished speakers lined up:

Alex Bateman will discuss the use of Wikipedia in RFAM, PFAM and the rest of science.
Alex Pico will tell us about the continued expansion of WikiPathways.
Andrew Su will be on tap to discuss the latest and greatest breakthroughs in the Gene Wiki.
Firas Khatib will tell us how his team managed to translate protein folding into a game (called Foldit) and that game into an expanding series of scientific discoveries.

With nearly 7 billion people on the planet and a rapidly increasing number of them connected to the Web, it is time to figure out how to work together effectively, both to solve the problems that we are collectively causing and to solve the problems whose solutions will extend and enrich all of our lives. Each of these talks will try to help answer the question of how we can translate the incredible scale, connectivity, and creativity of the world's population into scientific progress. Hope to see you there.

Semantic Publishing workshop at ESWC in Greece

2012-01-10T08:20:00.000-08:00

I'm helping to organize this exciting event, please consider submitting a manuscript and or attending. From the official call for papers:

http://sepublica.mywikipaper.org/SePublica2012 an ESWC2012 Workshop. May 27-31, Heraklion, Greece.
At Sepublica we want to explore the future of scholarly communication and scientific publishing. As we are going through a transition between print media and Web media, Sepublica aims to provide researchers with a venue in which this future can be shaped. Consider research publications: Data sets and code are essential elements of data intensive research, but these are absent when the research is recorded and preserved by way of a scholarly journal article. Or consider news reports: Governments increasingly make public sector information available on the Web, and reporters use it, but news reports very rarely contain fine-grained links to such data sources. At Sepublica we will discuss and present new ways of publishing, sharing, linking, and analyzing such scientific resources as well as reasoning over the data to discover new links and scientific insights.

Workshop Format

We are planning to have a full day workshop with two main sessions. During the first part of the workshop accepted papers will be presented; the second part of the workshop will address by means of focus groups two main questions, namely “what do we want the future of scholarly communication to be?” and “how could data be preserved and delivered in an interactive manner over scholarly communications?”. These focus groups will be followed by a panel discussion. As an outcome of these activities we will have a communique that will be the editorial for the workshop proceedings,

Dates

* workshop papers submission deadline: Feb 29
* workshop papers acceptance notification: April 1
* workshop papers camera ready: April 15

Submission

https://www.easychair.org/conferences/?conf=sepublica2012

Issues to be addressed
Representation:
Formal representations of scientific data; ontologies for scientific information
What ontologies do we need for representing structural elements in a document?
How can we capture the semantics of rhetorical structures in scholarly communication, and of hypotheses and scientific evidence?
Integration of quantitative and qualitative scientific information
How could RDF(a) and ontologies be used to represent the knowledge encoded in scientific documents and in general-interest media publications?
Connecting scientific publications with underlying research data sets
Technological Foundations:
Ontology-based visualization of scientific data
Provenance, quality, privacy and trust of scientific information
Linked Data for dissemination and archiving of research results, for collaboration and research networks, and for research assessment
How could we realize a paper with an API? How could we have a paper as a database, as a knowledge base?
How is the paper an interface, gateway, to the web of data? How could such and interface be delivered in a contextual manner?
Applications and Use Cases:
Case studies on linked science, i.e., astronomy, biology, environmental and socio-economic impacts of global warming, statistics, environmental monitoring, cultural heritage, etc.
Barriers to the acceptance of linked science solutions and strategies to address these
Legal, ethical and economic aspects of Linked Data in science

Scientific Games in Genome Biology

2012-01-02T07:00:00.000-08:00

Happy New year everyone! In case you are looking for some inspirational reading to start off 2012, Andrew and I wrote a Research Highlight for Genome Biology on Games With a Scientific Purpose. Kudos to Andrew for convincing them to make it open access so you can actually read it...

Research highlight Games with a scientific purpose Good BM, Su AI
Genome Biology 2011, 12:135 (28 December 2011)
[Abstract] [Full text] [PDF]

Gene Wiki on NAR database cover

2011-12-22T13:37:00.000-08:00

The Gene Wiki Rainbow just got a little bit more famous. Check it out on the cover of the 2012 Nucleic Acids Database issue. Thanks again Martin K.!

The Gene Wiki on the cover of NAR

Mining the Gene Wiki

2011-12-15T10:20:00.000-08:00

Our article about mining ontology-based gene annotations from the text of the Gene Wiki just came out at BMC Genomics. Yay!

In the article, we discuss the results of what I think might be the simplest text-mining strategy that could possibly work. Based on the premise that each Gene Wiki article is fundamentally about one particular gene, we make the simplifying assumption that all of the concepts detectable in the article are descriptors of what that gene does. With those assumptions in place, we use the NCBO annotator to detect concepts from the Gene Ontology (GO) and the Human Disease Ontology (DO) in the text of articles about genes. Each detected occurrence thus produces a candidate annotation for the gene. From the article:

For example, we identified the GO term ‘embryonic development (GO:0009790)’ in the text of the article on the DAX1 gene: “DAX1 controls the activity of certain genes in the cells that form these tissues during embryonic development”. From this occurrence, our system proposed the structured annotation ‘DAX1 participates in the biological process of embryonic development’. Following the same pattern, we found a potential annotation to the DO term ‘Congenital Adrenal Hypoplasia’ (DOID:10492) in the sentence: “Mutations in this gene result in both X-linked congenital adrenal hypoplasia and hypogonadotropic
hypogonadism”.

We found that, in terms of precision, this simple approach worked pretty well on detecting gene-disease annotations (90-93%) but not nearly as well at detecting gene-function (GO) annotations (48-64%). As you might expect, the recall equation worked in the opposite direction with many more potential GO annotations discovered (11,022) then DO annotations (2,983). Though there was some overlap, the majority of the predicted annotations did not have any match in existing annotation databases, showing that the Gene Wiki contains some knowledge that centralized resources like the Gene Ontology Annotation database do not yet represent and that basic text mining provides a way to access that knowledge computationally.

But, you say, that precision for the GO is really low, what use is this really? For applications that require 100% accuracy, like a curated database, well you would need to curate the predicted results and that might be quite a lot faster than searching through PubMed to find them all from scratch. As it turns out, there are also other kinds of applications that can take advantage of data like this that has noise in it. As long as there is a strong signal within the noise, probabilistic techniques, like enrichment analysis, can work. This is possible because, although many of the individual annotations might turn out to be incorrect, as a group they are far far from random.

For more details, read the paper ;).

Grain of Sand Quote - stolen!

2011-11-21T21:39:00.001-08:00

I was amused to see an article come out today that used the same introductory quote as we did six years ago. The paper from today is called "A world in a grain of sand: human history from genetic data" and is available from Genome Biology. I would tell you what its about, but neither I nor the Scripps Research Institute where I work has paid for access to it. You can however access our very cool paper called "Sand DNA-a genetic library of life at the water's edge" for free! Oh yes, the quote is:

‘To see a world in a grain of sand,

and heaven in a wild flower,

Hold Infinity in the palm of your hand,

and eternity in an hour.’

—William Blake from Auguries of Innocence

While I can't say which is the better manuscript, I'd wager that we win for most appropriate use of that quote in the intro.

Quotes from Reality is Broken

2011-11-18T14:36:00.001-08:00

I'm currently supposed to be writing an article about scientific discovery games in biology, but I have writer's block. So instead, I'm writing here.. which is much easier! The article I am not currently writing will discuss recent successes like "Algorithm discovery by protein folding game players" by Firas Khatib and others. In preparing to write this article (i.e. more not-writing), I assembled some inspiring quotes from the fantastic book "Reality is Broken" by Jane McGonigal. I share them here below because, well they made me think a little bit and perhaps they will do the same for some else, and this allows me to push back my real work by another 5 minutes..

“It is games that give us something to do when there is nothing to do. We thus call games “pastimes” and regard them as trifling fillers of the interstices of our lives. But they are much more important than that. They are clues to the future. And their serious cultivation now is perhaps our only salvation” - quote that opens McGonigal’s book - from Bernard Suits

“Games aren’t leading to the downfall of human civilization. They’re leading to its reinvention” (p354)

“Game developers know better than anyone else how to inspire extreme effort and reward hard work. They know how to facilitate cooperation and collaboration at previously unimaginable scales. and they are continuously innovating new ways to stick with harder challenges, for longer, and in much bigger groups.” (p13)

“Game design isn’t just technological craft. It’s a twenty-first-century way of thinking and leading. And gameplay isn’t just a pastime. Its a twenty-first-century way of working together to accomplish real change.” (p13).

“Anything else you think you know about games, forget it for now. All the good that comes out of games-every single way that games can make us happier in our everyday lives and helps us change the world-stems from their ability to organize us around a voluntary obstacle” (p34)

“Compared with games, reality is unproductive. Games give us clearer missions and more satisfying, hands-on work.” (p55)

“If you were able to focus the attention of the entire planet on a single goal, even just for one day, and even if it just involved dispatching aliens in a video game, it would be a truly awe-inspiring occasion. It would give the whole earth goose bumps.” (p112)

great article on Internet Science

2011-11-14T11:28:00.001-08:00

This is the best news article about science in the age of the Internet that I have ever read. If the title is at all interesting to you, its worth a read:

"How crowdsourcing is changing science"

by Gareth Cook.

Boston Globe, November 11, 2011

Gene Wiki article out today at NAR

2011-11-11T11:36:00.001-08:00

The articles for the annual database issue are starting to appear in the NAR collection. My favorite one this year is, immodestly perhaps, ours about the Gene Wiki! The simple message here is that the Gene Wiki is continuing to grow and that the content remains very high quality overall. For more information, the abstract is below, and of course the paper is freely accessible online:

"The Gene Wiki is an open-access and openly editable collection of Wikipedia articles about human genes. Initiated in 2008, it has grown to include articles about more than 10 000 genes that, collectively, contain more than 1.4 million words of gene-centric text with extensive citations back to the primary scientific literature. This growing body of useful, gene-centric content is the result of the work of thousands of individuals throughout the scientific community. Here, we describe recent improvements to the automated system that keeps the structured data presented on Gene Wiki articles in sync with the data from trusted primary databases. We also describe the expanding contents, editors and users of the Gene Wiki. Finally, we introduce a new automated system, called WikiTrust, which can effectively compute the quality of Wikipedia articles, including Gene Wiki articles, at the word level. All articles in the Gene Wiki can be freely accessed and edited at Wikipedia, and additional links and information can be found at the project's Wikipedia portal page:http://en.wikipedia.org/wiki/Portal:Gene_Wiki."

Stepping towards a Semantic Wikipedia

2011-10-08T00:00:00.000-07:00

It is now possible to specify the nature of the relationships between things described by Wikipedia articles directly in the context of the article. The image below is a screenshot taken a few moments ago of the Phospholamban article on Wikipedia (with excited arrows added). The infobox at the top right is dynamically generated from semantic markup in the article using a Wikipedia user script written by my colleague Sal and accessible from his user page.

Semantic markup now live in Wikipedia

How it works

Wikilinks in the article have been annotated with the kind of relationship that they indicate using the Semantic Wiki Link (SWL) template. The template allows any Wikipedia editor (including you!) to specify the type of connection that exists between the article where the link is being placed and the target of the link. This information is encoded following the microformat pattern. Essentially, we encode the meaning of the links in class attributes that wrap the link.

This works as follows:

Editor inserts a SWL into a Wikipedia article with this syntax:

{{SWL | target=protein kinase S | label=PKA | type =substrate_for}}
This means "the concept where you see this link is related to protein kinase A (labeled PKA) with the relationship type "substrate for". So, in the example above, it says: "Phospholamban is a substrate for PKA".
The {{}} denotes a Wikipedia template. Templates can take parameters (here parameters are separated by |'s) and use them to produce new WikiText dynamically which, in turn, is rendered as HTML when a page is loaded.

When the page is rendered, the template generates the following semi-semantic HTML markup (with some formatting omitted for clarity):

Programs, like the script that generated that infobox and added the green highlighting, can look for the SWL class attribute can then extract the meaning of the SWL links based on the class of its first child element - here "substrate_for".
In addition, when the template is processed it adds a category to the article it is placed on that corresponds to the relationship type. (See for example, the category for substrate.) This category provides a logical grouping (e.g. all things that serve as a biochemical substrate) but, perhaps more importantly, it provides a place to record the meaning of the relationship. This meaning can be defined as text, but can also be defined through reference to external sources such as ontologies on the semantic web.

Why its awesome

This pattern makes it possible for the vast number of Wikipedia users to simply and easily contribute machine readable content to the Web.  This enormous user community collaboratively created the world's largest encyclopedia and one of the most valuable websites on the planet.  Who better to help build the semantic Web?  While, technically, the microformat-like implementation leaves much to be desired in terms of its robustness and its precision, it is a solution that can work.  This is demonstrated by the success of projects like Google's recent recipe search that are based entirely on simple microformats.  

How you can help

This is a new idea that not everyone in Wikipedia will be thrilled to see.  They will claim that the SWLs will clutter the markup and will not provide enough value to make it worth it because Wikipedia itself does not support semantic links.  You can help by:

Using the template to enhance articles.
Writing code that makes use of the added meaning such as user scripts, aggregators, or scripts that import the relationships into other structured repositories like FreeBase or DBpedia.
Helping define the nature of the semantic links (at their associated category pages) and mapping them to properties defined in ontologies.
Discussing (and voting for) the idea on the various 'talk pages' on Wikipedia.

Why its awesome again

Did I mention that this pattern makes it possible for the vast number of Wikipedia users to simply and easily contribute machine readable content to the Web?  Thats pretty cool if you think about it...  

Update: To make the user script work for you so you can see the infobox, do this:

Create a Wikipedia user account if you don't have one already
Go to/create your user page. (e.g. my user name there is i9606 and my user page is located at http://en.wikipedia.org/wiki/User:I9606 )
Edit your user page add this to it -

[[/common.js]]

Visit your new common.js add this to it -

importScript('User:Sal9000/SWLinfobox.js');

When that is saved, you should be all set. Now go visit an enhanced page like Phospholamban and look for the green box at the upper right corner.

The script will run whenever you access a Wikipedia page while you are logged in to your account. Its a lot like GreaseMonkey, (which I've had some fun with in the past) but its not tied to your browser and will only work on Wikipedia. If enough people like a user script, it can be added to the default set of Wikipedia user preferences.. which would be pretty cool ;). The best part? You (or a programmer friend of yours) can write your own script and make it do whatever you want with the data!

Semantic Web Service Course in Vancouver

2011-08-16T13:45:00.000-07:00

My old friend and PhD advisor Mark Wilkinson is organizing an upcoming short course on semantic web services. Aside from a basic intro to the relevant pieces of the semantic web, you will become deeply ingrained in SADI. See below for details.

Training Courses in Web Publishing - Scientific Data and Services in Vancouver, BC, Canada.

Updated information!!

Registration Workshop
Until October 21st, 2011 October 22nd-23rd, 2011

"The Semantic Web is emerging as a new and more powerful standard for data and knowledge representation compared to the traditional Web. In this very unique, hands-on workshop you will learn how to use the Semantic Automated Discovery and Integration (SADI) framework to publish your data and analytical tools on the Semantic Web to make them easily accessible, and easier to integrate with other data and tools. Space is limited: don't get left behind in the Semantic paradigm shift! Register today at http://sadiframework.org/training/WS2/index.html."

The course provides theory and practical skills in the semantic technologies that underpin the development of semantic web services.The training course is divided into the following modules:

Semantic Web and DL Reasoning
SADI Semantic Web Services

The course provides an excellent learning opportunity and serves as a venue for the exchange of ideas among a highly interdisciplinary group of scientists. The intended audience for the course are:

Industry pioneers in workflow, grid and semantic technologies;
Web developers and researchers seeking to ensure the adoption of their technologies as part of a growing ecosystem of easily discoverable and accessible data and services;
Grid computing architects and developers interested in building globally interoperable systems;
Workflow system developers coordinating computational and data resources across the web using semantics;
Researchers involved in web-based knowledge discovery
The cost per participant is:
Regular Registration (375 CAN$)
Student Registration (325 CAN$)On site Registration (500 CAN$)

We ask that you broadcast this information widely to your colleagues. Our flyer can be found at http://sadiframework.org/training/WS2/WS2_flyer.pdf. Please do not hesitate to contact us at spaixao@unb.ca should you have any question.

GenomeWeb article on wiki mining mashup

2011-08-08T15:49:00.000-07:00

I guess we had a reporter in the audience at our talks at ISMB. Its always fun to get noticed though its a little scary to see something written about your work without ever having spoken with the author. In this case they got almost everything right. Thanks for the article Genome Web!

Gene Wiki Pulse tweets again!

2011-08-04T12:20:00.000-07:00

After a month long summer vacation, the Gene Wiki Pulse Twitter feed is once again alive and tweeting. While I suppose this little guy probably isn't going to get anyone tenure, its one of my favorite pet projects here at Gene Wiki Central and I'm very happy to have resurrected it. I love it because it provides a live, global view of the work being conducted on the Gene Wiki by thousands of people all around the world. My interest is largely in understanding and improving the knowledge assembly system running here as a whole, so the 'oooh look how much is happening! feeling' is enough for me to be an excited follower but I do hope that the updates are also useful to people that simply want to know more about their favorite genes...

For the technically inclined, the vacation was the result of the system clock on our server drifting forward about 10 minutes into the future (and ISMB and the move to the new office). Since the pulse only looks two minutes into the past, it suddenly became highly unlikely to pick up any new edits! Thanks to our new associate Erik for fixing the network time on our server and for suggesting the new, much more efficient watchlist-driven approach to monitoring Gene Wiki activity that is running now. If you want to write your own Wikipedia watching and tweeting program you can find the source code for this in the gene wiki repository in the class StreamAccess.

bio-ontologies 2011 - ontologies need lexicons

2011-07-22T11:11:00.000-07:00

I recently returned from my first ISMB in Vienna, Austria where I presented a SNPedia-Gene Wiki mashup at the Bio-Ontologies special interest group meeting. (You can get to most of the other papers at the Knowledge Blog). In quite unusual and positive fashion, the SIG came to a close with a discussion that I, at least, felt was a specific call to arms that may end up seeding a number of research programs for the next little while. In a nutshell, we need ontologies to provide better support for text-mining tools by providing examples of the terms/phrases that best match the concepts in the ontology as they occur in important, relevant bodies of text like PubMed abstracts. Read on for the non-nutshell explanation..

There were a number of papers in the SIG that made use of text mining services that performed concept recognition using biomedical ontologies for one reason or another. It seemed like every other talk mentioned the NCBO Annotator, including ours. Basically, despite the best efforts of the semantic web community and the world's databases, the great majority of the knowledge in the world is still shared as plain old text. As a result, if you want to get something useful done immediately, like test for an association between a drug (like Viox) and a nasty outcome (like a heart attack), you can do a much better job of it if you can somehow compute with large bodies of text (in this case clinical records) than if you limit yourself to the comparatively small body of knowledge that has been clearly represented database records (see [1]).

Now, to 'compute with large bodies of text', most approaches in the biomedical domain involve some mapping from the text to terms in an ontology. Such mappings make it possible to test for associations between the mapped concepts (e.g. drugs and diseases) using fairly straightforward statistical techniques. The trouble of course is that the mapping process generally remains highly error-prone (*). One of the reasons for this is that text2concept services like the Annotator are actually text2text services.

To find an occurrence of the ontology term GO:0006915 in text, the simplest, fastest way to go about it is to search for its preferred label 'apoptosis' in the input text. This is basically all the Annotator does (with some tricks to make it go really fast and to catch non-exact matches). This means that any concept matching system like the Annotator is fundamentally dependent on the labels associated with the concepts in the ontologies that are fed to it. If the labels don't line up well with actual usage in the texts where you want to perform the mapping, than the system doesn't work very well. We noticed this in the context of our work on identifying ontology terms in the text of Gene Wiki articles.

In the Gene Wiki, terms from some ontologies are much easier to match than others. For example, we found that the Annotator picked up Disease Ontology terms with about 90% precision while its precision with the Gene Ontology was less than 50% on exactly the same text. Others have reported similar results (though in many cases the work is not published because the scores are so low). In fact, the text mining community seems to have basically given up on the challenge of mapping to the GO. After the final keynote at the SIG, I asked Martin Krallinger if the BioCreative text mining competition was planning to work on the GO again as it had in its first iteration several years ago and he basically said no because it was impossible..

I suggest that a key reason that the GO isn't working very well with the Annotator and other tools is that the labels attached to concepts weren't generated with text mining in mind. As a result, they sometimes never occur at all in the typical target text (journal articles and abstracts in PubMed) and sometimes occur over and over again but mean something completely different than the concept they are linked to in the ontology.

To make advances in our use of the GO in the context of of text mining we need more useful labels for the concepts. I suspect this is the case for many of the ontologies in use on the semantic web. So, here are a couple of IMO very achievable things that our community needs done.

1) The semantic web community as a whole would benefit from a consistent, clear RDF structure for capturing textual representations of ontology concepts in OWL/RDF that are intended specifically for use by concept detection services. An RDF property designed to signal text mining tools that its target string was inserted specifically for them would make it possible for ontology developers to intentionally aid those who build concept recognition services. RDF:label does not solve this problem because the label that the ontology developers might want to see for the concept might not be the same as the string text miners need because the contexts are different. This might seem like too much work for ontology developers and that the NLP people should really be doing this kind of work but I think its the most feasible. If you consider the Annotator problem - to build a concept recognition service for any ontology submitted to the BioPortal collection (now over 200 ontologies) - then some level of standardization is required.

2) The biomedical community needs a better lexicon for the Gene Ontology. Though this is a general problem for any ontology that could be used in a text mining context, the GO is a special case that needs to be dealt with regardless of whether or not we solve the problem for other ontologies. Even if we have a one off, GO-specific invention, we really need to be able to figure out when the concepts from the GO occur in text and to simply throw up our hands and say that 'the GO was not designed for text mining' is a pointless cop out. This can and should be done.

----

* But note that we are seeing that, with enough text, the underlying signal can sometimes drown out the errors from the NLP. (See the unreasonable effectiveness of data for more on this topic at even grander scales).

[1] "Annotation Analysis for Testing Drug Safety Signals" by Paea LePendu et al from bio-ontologies this year. (their paper isn't in the knowledge blog for some reason, but I'm sure Paea will send it to you if you ask him)

Gene Wiki Rainbow

2011-07-11T10:23:00.000-07:00

Last summer I posted an image of the Gene Wiki hyperlink network, aptly titled "the gene wiki hairball". The image was picked up by noted artist/scientist Martin Krzywinski and used as an example of why hairballs are a terrible visualization. Perhaps out of guilt for making an example out of us and/or perhaps out of interest he has helped us improve our thinking about how to visualize networks substantially. Here is a Circos view of the top 100 genes in the gene wiki, the editors that created the articles, and the diseases and compounds that the genes are linked to. It will be presented as a guerilla poster[1] at ISMB this year so please stop by and have a closer look!

Aside from bringing an artistic aesthetic to scientific illustrations, one of Martin's main contributions (IHOP) is that he understands and uses space to convey meaning effectively. In a hairball, the only consideration of space in the layout algorithm is to reveal as many nodes as possible. In a Circos diagram, place can be attached to semantics. This fundamental idea is used to an even greater extent in Martin's latest layout invention, Hive Plots. Watch out for the Gene Wiki Hive Mind visualization - or better yet, write to us and help us build it!

[1] guer·ril·la post·er /gəˈrilə/ /ˈpōstər/

Noun: An uninvited poster displayed at a scientific conference.

why research blogging has mattered to me

2011-05-29T22:47:00.000-07:00

A little while back I had the honor of speaking on a 'careers in bioinformatics' panel put on by my former graduate program at the University of British Columbia. Other panelists included Inanc Birol, Bioinformatics Group Leader, Michael Smith Genome Sciences Centre (who good-naturedly called postdocs the 'burger flippers' of science), Phil Hieter, Professor of Medical Genetics, University of British Columbia, and Anthony Fejes, Co-founder, Zymeworks and PhD candidate at UBC. Blogging came up a few times in the discussion as Anthony and I have both been at it for a while now and have seen a clear positive impact on our careers from it. Anthony's blog has become so popular that he has started to get invited to conferences solely because the organizers want him to blog about them.. My blog hasn't achieved the audience that fejes.ca enjoys but it has managed to connect me with some excellent people that I wouldn't normally have met and through them with opportunities I never would have discovered. I think its worth jotting down a couple examples.

Blogging got me to Sci Foo. While its hard to say exactly why you get invited I have a pretty solid guess that my writing here about semantic social tagging is what caught the attention of Timo Hannay and resulted in my trip to the GooglePlex. Had I waited for my work on tagging to be published through a traditional journal this never would have happened.

Blogging connected me with Barend Mons and what would become the Concept Web Alliance. Barend picked up on the same set of posts (mainly the big one that turned into a lovely rejected paper) and eventually invited me to do a postdoc in his group. Babies and geography aside I would probably be working with him now. Though I didn't end up working with him in Leiden, he did invite me to the first Concept Web Alliance meeting where I met Andrew Su, who would become my employer nearly a year later..

Blogging got me a better Gene Wiki network visualization! A while ago I posted a Gene Wiki hairball picture illustrating the hyperlink network emerging from the articles in the gene wiki. This post was picked up and used as an example of what not to do when visualizing large networks by Martin Krzywinski - author of the popular Circos visualization system and the up and coming hive plot paradigm. After discovering our ahem 'connection', we've been working with Martin to build something more beautiful and more informative. Posts to come about that one..

In short, sharing ideas openly here in i9606 has been one of the best parts of my experience in science and one I hope to continue for many years to come.

moving to tSRI

2011-05-25T21:44:00.000-07:00

The Gene Wiki and I are following Andrew Su down the street to the nascent SuLab at the Scripps Research Institute. Its not going to be an enormous change for me - its literally the next building down the street, I'll remain a postdoc, and I will continue with ongoing Gene Wiki projects - but I think its still going to be a very positive move and I'm excited about it. Scripps is a much more academic environment than GNF and I think that mentality will suit me and my projects better. One thing I'm going to strive to do at my new home will be to increase my blogging activity - more on the how and why of that later..

Integrating the Gene Wiki with traditional publishing?

2011-05-06T16:42:00.000-07:00

While we (Andrew Su and I) like to talk about the successes of the Gene Wiki - articles like the one for Reelin that represent arguably the best consolidated body of text associated with the gene - there remain some rather glaring holes in its content. A couple months ago I had a look for under-developed articles linked to genes with extensive numbers of publications. With a small bit of hacking I uncovered a list of 2,553 genes that were linked directly to more than 20 PubMed citations (using NCBI's gene2pubmed) but had less than 100 words of text in their Gene Wiki article. (Up to the previous period, this post contained 105 words.) From this list I found 151 genes with more than 100 PubMed citations and less than 100 words of wiki text.

An example is the PIN1 gene. When the analysis was run, this gene was linked to 154 citations in PubMed yet had only 2 sentences in the Gene Wiki. So... how do we fill in these gaps? This is, of course, the fundamental question associated with wikis or any other attempt to harness community intelligence and there is no easy answer. One model that we are very interested in was pioneered by Alex Bateman and colleagues at the journal of RNA Biology. When hopeful authors submit an article about a new RNA family to the journal, it is a condition of publication that they contribute an article to Wikipedia about that family. Aside from being a generally good thing to do as far as sharing knowledge with the world, these articles are subsequently used to manage the annotations for RNA families in the Rfam database (e.g. snoZ107_R87). After a few years of operation, the Rfam team published an article that, among others things, celebrated the success of the Wikipedia connection. So, how might we expand upon this model to tackle the challenges facing the Gene Wiki?

The beauty of this approach is that it does not rely on any changes to the incentive system currently operational in science. Scientists need to publish in peer-reviewed journals. Rather than complaining about the inefficiency of this outdated process and suggesting social changes with no obvious way to achieve them, lets see what we can do to make the system work for us as it stands. Lets create a way for scientists to obtain real publications in real journals and have Gene Wiki article content generated as a natural part of the process. Here is one idea.

A Gene Wiki Meta-Journal Special Edition

In this model we would work with a number of smallish, topic-focused journals to requisition short review articles about the molecular function and phenotypic relevance of individual human genes. (By phenotypic relevance I mean the connection between the gene and something that non-scientists might care about such as a role in a disease or a connection to some human attribute such as height, hair color or athletic performance.) These review articles would be published in journals appropriately matched to the key phenotypes associated with the gene. For example, we might imagine requesting a review article in the Journal of Investigative Dermatology about the gene Filaggrin because the most important variations in this gene have been shown to relate to skin conditions such as eczema. Each of these phenotypically targeted gene review articles would be linked from and would link back to a central article that described the meta-journal concept - ideally in a journal with an audience broad enough to span each of the more niche-specific journals that participated in the experiment. Following the RNA Biology model, a condition of publication would be to update or create the relevant Gene Wiki article with content from the submitted review article.

While logistically challenging, this approach appeals to me because it continues with the theme of tapping into the 'Long Tail'. If we can distribute the labor out among a larger number of journals we ought to be able to connect with a larger number of individual contributors. In addition, it might be appealing to the editors and contributors to more niche-specific journals to participate in a project with broad visibility. As an alternative, we might consider attempting to organize a gene-focused special edition similar to the annual Nucleic Acids Research database edition in one gene-focused journal (e.g. Genome Biology), but it seems unlikely that this approach would have the same potential breadth of impact. Also, following a phenotypic rather than molecular orientation aligns well with Wikipedia's notability criterion and hence might help to generate article content that would meet with less resistance from current Wikipedia editors.

If you have any thoughts on this idea (or if you have better ideas!) I would love to hear from you.

Collaborative Innovation in Philadelphia

2011-04-05T20:33:00.000-07:00

Today I presented a high-level view of the Gene Wiki project to a group of people interested in expanding the application of "collaborative innovation" in the biomedical domain. My presentation (and my lack of a suit) definitely felt like an outlier in this group. Many of the other talks brought up the apparently looming demise of the pharmaceutical industry due to a lack of innovation and made suggestions about how companies and consortia could make changes that would allow them to tap into much broader, more diverse intellectual and physical resources using a variety of mechanisms. We also had presentations about how the same kinds of cost-saving, creativity-enhancing approaches might be used to tackle neglected diseases (aka diseases that are not likely to make the pharma companies huge amounts of money by curing as opposed to say, cancer). The special sauce here is this "collaborative innovation" stuff. This seems to some extent to reiterate the age-old solution to complexity - 'divide and conquer'. The problems of developing novel drugs today are diverse - any successful solution is the result of the execution of many many different tasks. The key challenges for people in this industry (and many others I imagine) seem to be knowing which of the tasks they are very good at completing, which tasks they are not, who the people are that can complete these tasks, and how to legally transact mutually beneficial collaborations with those other people.

For more information see the conference website for abstracts and links to the talks.

Random notes to self:

I love talking about the gene wiki! Its just a great story to tell. I'm really looking forward to the time that my part of this story is half as interesting as the parts that have already transpired.
Why do I spend so much time worrying about and getting ready for presentations that are largely gone the moment they are completed and so little time on these blog posts that tend to stick around for years?
This evening I spent far too much time preaching that games were the future of this domain and that Jane McDonigal had the keys to saving the world in her book 'Reality is Broken'. I couldn't help it, I got excited!
I should have spent more time listening to what people like Hassan had to say - especially as he seems to have been studying game-like systems for large-scale collaborative innovation for a long time...

Chafing at censorship

2011-03-31T15:59:00.000-07:00

Big brother is watching me and he says I can't work on the GeneWikiPulse today...

The Blue Coat Web Filter in action

This is one of the downsides of working for a larg(ish) company with a "protective" approach to information technology. The big upsides of course are the salary and the stock options. Oh, wait... I'm a postdoc...

TBI-AMIA 2011 Brain Dump

2011-03-10T12:44:00.000-08:00

Here is a very rapid dump of a few of the tidbits that stuck in my brain after spending the last few days at the Translational Bioinformatics (TBI) conference for 2011.

The main recurring themes of interest for the year were:

Social networking for scientists: dirrecttoexperts, vivo, open social, patient recruitment (e.g. army of women). All efforts are heavily into the linked open data concept.

Temporal reasoning for clinical data: the chronology of the patient's symptoms is an important and underrepresented aspect to mine/model in patient records

Naive Bayes : came up at least twice in important places - once in Lincoln Stein's keynote (used in the automatic expansion of the Reactome database) and once in the nominated best paper by Wei Wei "The Application of Naive Bayes Model Averaging to Predict Alzheimer’s Disease from Genome-Wide Data". Its an effective method to integrate many many different sources of evidence into a single predictor - a relatively simple machine learning algorithm that scales up well with a lot of data.

Genomic complexity : keynotes about cancer and microbiome highlighted the incredible diversity of genomes. One example described the difference between a tumor cell and a normal cell one inch away on the order of 50,000 SNP differences - add on top of that genomic rearrangements and epigenetics and well... its complicated. Complete genome sequencing renders all diseases orphans from a drug development perspective - major changes needed for pharma... Major opportunities as well.

Politics : NCBO vs. OBO vs. UMLS

Each of these groups is or will be competing for the same pot of money to solve very similar problems. While members do collaborate, e.g. key UMLS players advise NCBO, they have serious disagreements about how things should be done and I noticed some pretty emotional discussions - mainly related to fears caused by expected funding cuts.

name	goal	conflict with OBO	conflict with UMLS
NCBO	empower researchers with access to ontologies and related tools	NCBO is very open about including ontologies. OBO is not - they believe strongly in their foundry principles (and their distinctive file format) and it annoys them that anyone can get their ontology into NCBO.	See goals.. and think $$ NCBO has a different approach to concept detection that annoys the MetaMap people. MetaMap feels like NCBO is throwing away years of work on NLP that should be used.
OBO	provide "high quality" collection of biomedical ontologies		Some OBO members seem to have a longstanding dislike of the way that the UMLS is modeled. OBO believes they are right and the UMLS is wrong and have effectively ignored its presence in the development of their ontologies.
UMLS	empower researchers with access to ontologies and related tools

Unsurprising : lots of people building ontologies to use for structuring data and lots of people applying text mining approaches to mine unstructured data

There was a whole lot more than that, for some more info, have a look at the conference papers, and at Russ Altman's year in review.

with a flower in my hair

2011-03-06T12:54:00.000-08:00

Assuming the fog clears enough to land, I'm heading to San Francisco today to attend the 2011 AMIA Summit on Translational Bioinformatics. I'll be presenting Tuesday morning about mining structured gene annotations from the text of the Gene Wiki. Supporters and hecklers would be welcome!

Gene Wiki on Twitter

2011-02-28T20:45:00.000-08:00

As you may notice in the new live twitter feed to the right (if you are reading this at blogger), a new entity has entered the twittersphere. The GeneWikiPulse provides a live (well within a couple minutes) feed of editing activity on the Gene Wiki. In a nutshell, a program is watching all the articles in the Gene Wiki and doing its level best to turn every editing event into an informative message in 140 characters or less. Each such tweet is signed with a hashtag for the official gene symbol and the responsible editor.

This is the first little step we've taken to increase community engagement for Gene Wiki since I got started and I'm very curious to see how people react. If you have any ideas for improving this stream, let me know!

For a tad more info., check out the post describing the Gene Wiki Pulse on the BioGPS blog.

Oh yes, and please go ahead and follow the GeneWikiPulse!

testing twitterfeed

2011-02-22T08:08:00.000-08:00

Just seeing if this works.. (it reads RSS feeds and tweets them).

TripleMap

2011-02-03T13:30:00.000-08:00

I had fun playing with TripleMap today. Its an RDF-backed Adobe Flash (Flex I assume) Web application for exploring graphical biomedical data.

Rapid conclusions

Its pretty - the visualization system worked well for my explorations
It has a few bugs (but it hasn't even reached a beta release yet so thats to be expected)
You are going to want to get more data into it as soon as you start using it. While there is enough in there today to get you started, anyone familiar with bioinformatics will notice that a lot is still missing (but I know more is already on its way).
Its definitely something to keep an eye on as it develops.

cash for semantic publishing research

2011-02-01T07:43:00.000-08:00

Following from my last post, here is an update from the SePublica workshop

SUBMISSION DEADLINE February 28

ELSEVIER BEST SEMANTIC PAPER AWARD

The Best Paper Award is presented to the author(s) deemed to have written the paper covering the most innovative and feasible proposal concerning semantic publishing in the workshop. All submissions to the SePuBlica workshop will be considered, and a panel of experts will rate the papers according to originality of the idea, feasibility and presentation. The Best Paper award is sponsored by Elsevier as an incentive for researchers working on defining the next generation of scientific publishing concepts. The Best Paper Award will be handed out at the end of the SePuBlica workshop.

As a cash prize, the Best Paper Award will receive: US$ 750

The runner-up will be awarded a prize of US$ 250.

CfP Semantic Publishing

2011-01-12T20:19:00.000-08:00

Minoan rhyton from Crete!

I'm on the program committee for this conference workshop so clearly you should submit something (and its in Crete!). See the call for papers below:

-----------------------------------------------------------------------------------------------------------------

1st International Workshop on Semantic Publication (SePublica 2011)

http://sepublica.mywikipaper.org

at the 8th Extended Semantic Web Conference (ESWC 2011)

http://www.eswc2011.org

May 29th or 30th, Hersonissos, Crete, Greece

Keynote by Steve Pettifer, Manchester University, UK.

“Utopia Documents and The Semantic Biochemical Journal experiment”

SUBMISSION DEADLINE February 28

The MISSION of the SePublica workshop is to bring together researchers

and practitioners dealing with different aspects of Semantic

Technologies in the Publishing Industry. How is the Semantic Web

impacting the publishing industry? How is our experience of

publications changing because of Semantic Web technologies being

applied to the publishing industry?

The CHALLENGE of the Semantic Web is to allow the Web to move from a

dissemination platform to an interactive platform for networked

information. The Semantic Web promises to “fundamentally change our

experience of the Web”.

In spite of improvements in the distribution, accessibility and

retrieval of information, little has changed in the publishing

industry so far. The Web has succeeded as a dissemination platform for

scientific and non-scientific papers, news, and communication in

general; however, most of that information remains locked up in

discrete documents, which are poorly interconnected to one another and

to the Web.

The connectivity tissues provided by RDF technology and the Social Web

have barely made an impact on scientific communication nor on ebook

publishing, neither on the format of publications, nor on repositories

and digital libraries. The worst problem is in accessing and reusing

the computable data which the literature represents and describes.

• Consider research publications: Data sets and code are essential

elements of data intensive research, but these are absent when the

research is recorded and preserved in perpetuity by way of a scholarly

journal article.

• Or consider news reports: Governments increasingly make public

sector information available on the Web, and reporters use it, but

news reports very rarely contain fine-grained links to such data

sources.

QUESTIONS AND TOPICS OF INTEREST

• What does a network of truly interconnected papers look like?

How could interoperability across documents be enabled?

• How could concept-centric social networks emerge?

• Are blogs and wikis new means for scholarly communication?

• What lessons can be learned from humanities and social science publishers

(i.e. going beyond scientific publishing towards scholarly publishing)?

• How could we move beyond the PDF?

How can we embed and link semantics in EPUB and other e-book formats?

• How are digital libraries related to semantic e-science?

What is the relationship between a paper and its digital library?

• How could we realize a paper with an API?

How could we have a paper as a database, as a knowledge base?

• How is the paper an interface, gateway, to the web of data?

How could such and interface be delivered in a contextual manner?

• How could RDF(a) and ontologies be used to represent the knowledge encoded

in scientific documents and in general-interest media publications?

• What ontologies do we need for representing structural elements in a

document?

• How can we capture the semantics of rhetorical structures in

scholarly communication, and of hypotheses and scientific evidence?

AUDIENCE

• researchers from diverse backgrounds such as argumentative

structures, scholarly communication, multi-modality in publications,

digital libraries, semantics in publications, and ontology

engineers.

• practitioners active in the publishing industry, repositories of

experimental information and document standards.

IMPORTANT DATES

Paper/Demo Submission Deadline: February 28, 23:59 Hawaii Time

Acceptance Notification: April 1

Camera Ready Version: April 15

SePublica Workshop: May 29 or May 30 (to be announced)

SUBMISSION AND PROCEEDINGS

Research papers are limited to 12 pages and position papers to 5

pages. For system descriptions, a 5 page paper should be

submitted. All papers and system descriptions should be formatted

according to the LNCS format

http://www.springer.com/computer/lncs?SGWID=0-164-6-793341-0

We encourage the submission of semantic documents. LaTeX documents in

the LNCS format can, e.g., be annotated using SALT

(http://salt.semanticauthoring.org) or sTeX

(http://trac.kwarc.info/sTeX/). We also invite submissions in

XHTML+RDFa or in the format or YOUR semantic publishing tool.

However, to ensure a fair review procedure, authors must additionally

export them to PDF. For submissions that are not in the LNCS PDF

format, 400 words count as one page. Submissions that exceed the page

limit will be rejected without review.

Depending on the number and quality of submissions, authors might

be invited to present their papers during a poster session.

Please submit your paper via EasyChair at

http://www.easychair.org/conferences/?conf=sepublica2011

The author list does not need to be anonymized, as we do not have a

double-blind review process in place.

Submissions will be peer reviewed by three independent

reviewers. Accepted papers have to be presented at the workshop

(requires registering for the ESWC conference and the workshop) and

will be included in the workshop proceedings that are published online

at CEUR-WS.

PROGRAM COMMITTEE

• Robert Stevens, Manchester University, UK

• Benjamin Good, GNF, USA

• Michael Kohlhase, Jacobs University, Germany

• Oscar Corcho, Politecnica de Madrid, Spain

• Steve Pettifer, Manchester University, UK

• Jodi Schneider, DERI, NUI Galway, Ireland

• Sebastian Kruk, knowledgehives.com, Poland

• Henrik Eriksson, Linköping University, Sweden

• Dagobert Soergel, University of Maryland, USA

• Tim Clark, Harvard Medical School, USA

• Paolo Ciccarese, Harvard Medical School, USA

ORGANIZING COMMITTEE

• Alexander García Castro, University of Bremen, Germany

• Christoph Lange, Jacobs University Bremen, Germany

• Anita de Waard, Elsevier, USA/Netherlands

• Evan Sandhaus, New York Times, USA

QUESTIONS? → sepublica@googlegroups.com

About the Code Delusion

2010-12-21T09:34:00.000-08:00

After randomly stumbling across a link to the article“Getting over the Code Delusion” in a link posted by Deepak Sing a couple months ago I’ve just spent the last hour readingit instead of getting my own work done and now I’m spending more time writingabout it. Sorry boss, but I couldn't help myself...

The central point that the article makes with surprisinglyelegant prose is that it is impossible to understand how life works entirelythrough the lens of a linear DNA sequence. Put very simply, life is much more complicated than we had hoped – readhis piece to get a better understanding of some of the mechanisms of thatcomplexity.

While I generally found the piece informative, enlightening and a pleasure to read, one elementthat I found unnecessary was a thread of what seem like bizarrely placed callsto some higher power. For example,

“When you encounter the meaningful, directed, andwell-shaped movements of a dance, it’s hard to ignore the active principle —some would say the agency or being — coordinating the movements.”

And close after,

“Seeminglyin the grip of the encircling DNA with its relatively fixed and stablestructure, yet responsive to the varying flow of life around it, the nucleosomeholds the balance between gene and context — a task requiring flexibility, a ‘sense’of appropriate rhythm, and perhaps we could even say ‘grace.’”

These, perhaps unscientific, references seem to arise in the text because thecomplexity of the totality of the factors that affect how cells and organismsfunction appears too great to hope to understand in a reductive way. Along the same lines he suggests:

"The search for precise explanatory mechanisms and codesleads us along a path of least resistance toward the reduction of understanding."

Now what does that mean for a scientist? How can we increase understanding without searching for precise explanations?

It puts me in mind of the unlimited complexity produced by simple underlying rules in Wolfram's "A new kind of science" - would love to hear what it puts in your mind.

Digital Science launches

2010-12-07T09:31:00.000-08:00

Digital Science, a "sibling" of Nature Publishing under the parentage of Macmillan Publishers looks like it could be an interesting new company in the scientific publishing domain. Interesting because it looks like they are really setting themselves up as a software company and perhaps more interesting because they seem to be doing so primarily through partnerships. Seems like it could be grand opportunity if you are trying to get the ball rolling with a scientific software company.

NCBO Annotator versus MetaMap on GO concept detection

2010-12-02T12:08:00.000-08:00

For the past little while I've been working on extracting candidate Gene Ontology annotations from the hypertext of articles in the Gene Wiki. In this work I have been using two of the premier tools for concept recognition in bioinformatics - MetaMap from the National Library of Medicine and the Annotator from the National Center for Biomedical Ontology. As elaborated on in this article, these systems work very differently and, depending on the input text, can yield substantially different results. Since I was at a loss when I first had to decide which tool to use, I thought I'd share some of the results of my experiments in the hopes that I might help some one else along in their decisions.

Input: text from about 10,000 Gene Wiki articles (both complete sentences and the titles of pages linked to from a gene page)
Output: concepts from the GO with some form of linguistic match to the text.

Current Rounded Results:
GO concepts detected: Annotator about 20,000, MetaMap about 35,000, Intersection about 14,000 (See diagram below)

In this case MetaMap is the clear winner in terms of recall. Based on some casual manual inspections and some less casual comparisons to other GO annotation sources, the Annotator earns a very slight edge in terms of precision (it produces slightly fewer false positives). For my application, I'm pooling the results of both tools.

Another important consideration is speed of execution. For my experiments, a locally installed version of MetaMap took about twice the time to run the same jobs as the Annotator Web service - despite the lag from network latency for the Annotator. The output from both tools was fairly easy to parse. (I found it much easier to simply parse the results myself than to work in the context of UIMA wrappers which are available for both systems.)

So there you go, same same but different.

Shakey on the Web

2010-11-22T08:40:00.000-08:00

Mike Bergman has a nice post about the lack of appropriate means for representing ambiguity on the Web that is worth a look.

It reminded me again of Shakey the robot - its promise and its ultimate failure. Shakey was one of the first 'autonomous' robots; capable of making its own decisions to guide its actions based on data from its sensors. In its heyday in the 60's and 70's it stirred up Jetsonian imaginations of robot vacuum cleaners and so forth but in the end, finally failed to deliver anything of the kind. Shakey's mind was trapped in a strict, orderly world of logical rules. Much like the character Brooks Hatlen from the Shawshank Redemption, when faced with the uncertain reality of the our world he simply couldn't cope. (Watch him shake in trepidation!)

The robot minds that did end up making it into our vacuum cleaners are based on very different kinds of 'minds'. In a movement initiated by Rod Brooks in the late 80's, now personified at home by the iRobot vaccum cleaners, in the academy by the field of 'probabilistic robotics' and on the podium by Stanley, the rules and central planning centers of good old fashioned AI are gone. In their place, we find statistical models that - like the world they were built to deal with and perhaps a bit more like human minds - constantly change to reflect the uncertainty of reality.

I fear that our current approach to "reasoning" on the semantic web bears much more in common with old Shakey than it does with Stanley.

Presenting at San Diego Semantic Web Meetup

2010-11-08T10:07:00.000-08:00

In a few weeks, I will be giving a seminar for the San Diego Semantic Web Meetup group. Its going to cover a fair amount of the same material we covered at the NCBO webinar, but you should come anyway because a) I have new data and b) I'm going to be there live (so you can bring your rotten tomatoes!).

** Update, I am now excited to be presenting with Denny Vandrecic, an expert on ontology evaluation and one of the founders of the Semantic Media Wiki.

How many articles in PubMed contain information about genes?

2010-10-16T14:57:00.000-07:00

In Figure 1 from this article about the Open DMAP information extraction system the authors suggest that, as of 2007, about 40% of the articles indexed in PubMed contained some information about genes that could be extracted from their abstract. That seems like a lot.. If anyone has any data to corroborate or to disprove that statement, please let me know.

--Correction. The article said 40% of recent articles had a gene mention. See comments from Joachim below that suggest a better estimate overall of 19% for articles with abstracts.--

Gene Wiki talk saved at NCBO

2010-10-07T11:04:00.000-07:00

A webex recording of our webinar "The Gene Wiki: Cultivating and mining community intelligence in biology" is now available on the NCBO website. It was a little weird talking to myself - in particular it was totally impossible to know if any attempt at humor was successful.. Overall I think it went pretty well though since we had a good number of relevant questions from the audience. I imagine it would probably be a good idea from a learning and improving perspective but I have yet to bring myself to actually listen to myself speak.. Thankfully the webcam was not in effect so I'll never have to watch myself. (Andrew had the first half and I had the second.)

NCBO webinar on Gene Wiki

2010-10-01T09:50:00.000-07:00

Andrew Su and I will be presenting "The Gene Wiki: Cultivating and mining community intelligence in biology" next week live and on the Internet!

Feel free to join in if you'd like to hear what we're up to. October 6 at 10AM Pacific time.

The first half (presented by Andrew) will give an overview of what the Gene Wiki is, why it was built, and how it is doing. The second part (by me) will describe the results of efforts to extract structured data from the text and the links of the gene wiki. (Warning, my part is very much a work in progress...)

community annotation workshop - what is working? what is not?

2010-09-30T08:21:00.000-07:00

If you are interested in progress on the front of community intelligence in biology, the wrap up from the GMOD community annotation satellite meeting is well worth a look.

A couple highlights

Fail: "UniProt, a pan-biology resource if ever there was one, had 46 million page views lead to only 9 comments being submitted. "
Win: Following in line with the "distributed grid of undergraduate students" approach to gene annotation, the CACAO project has achieved "stunning" success. They get teams of students to compete against each other on both quantity and quality of annotations. Quality is determined in Scrabble-esque fashion where points are awarded based on the opposing team's ability to successfully challenge and disprove the other team's annotations.

Trials and Tribulations with the UIMA wrapper for the NCBO annotator

2010-09-22T15:31:00.000-07:00

As I've recently learned, UIMA stands for 'Unstructured Information Management Architecture'. UIMA emerged from the bowels of IBM Research and is now a full-fledged, open source Apache software project. From the Apache description:

"Unstructured Information Management applications are software systems that analyze large volumes of unstructured information in order to discover knowledge that is relevant to an end user. An example UIM application might ingest plain text and identify entities, such as persons, places, organizations; or relations, such as works-for or located-at."

I came across UIMA via this article about "A UIMA wrapper for the NCBO annotator" by Christopher Roeder and friends from Colorado and Stanford. For those that are unfamiliar, the annotator is a fairly newish web service for identifying terms from biomedical ontologies in text. (Here is a nice little interface you can use to see what it can do.) As I'm always looking for ways to avoid reinventing the wheel, I was hoping I would be able to use this wrapper on top of this well established framework to quickly build up a nice client for processing some Gene Wiki-related text. It turned out that, aside from my hopes for a quick solution... this is pretty much what I found.

Your results may vary, but here are my key early experiences with the UIMA wrapper for the NCBO annotator :

I started from the BioNLP code available from the folks in Colorado.
I had a lot of trouble building it.

There were missing jar files referred to in the ant build file.
Once everything was there I could not run the main application via ant and quickly gave up on ant and started to rebuild the entire project and its dependencies in Eclipse. This was probably a mistake - I should have stuck with the provided ant-based setup and figured it out but I got impatient...
Once you get into the insides of the system you will see that it is quite complex.. If you are not an advanced Java programmer you should not look inside.. - I thought I was fairly advanced until I got started.
The saving grace throughout this process was that the primary author on the wrapper paper - Chris - , was very responsive, patient and helpful.

Once built, it again took me a while to understand how the whole thing was supposed to work.

A key step forward was launching the 'CPE' GUI application - otherwise known as the 'Collection Processing Engine Configurator'. (main class was org.apache.uima.tools.cpm.CpmFrame)
From here I finally started to find the 'plug and play' kind of functionality I was hoping for in this framework. From the GUI, you can choose from XML files that configure components of an analysis such as a directory reader, a sentence detector, the NCBO annotator service(!!), and a basic results exporter. These XML files appear in a 'desc' directory that comes with the distribution. Each one maps to a class containing the code they refer to that uses the parameters they contain/collect.

Now I've got it running (this took a few days) - here are the main conclusions

If you aren't well-versed in Java, don't try this at home
This was definitely slower than writing my own client, BUT!!
This client is probably much better than what I would have naively done because:

it includes error handling
it throttles itself via sentence splitting (enabled via a third-party UIMA component)
it seems to go much faster.. I'm not sure why (perhaps parallel requests via multi-threading, perhaps the sentence splitting step makes the annotator happier)
it has also pushed me much closer to being able to run a large collection of powerful tools such as Open DMAP (whether I ever head down that road, time will tell).

In conclusion, it feels like this project (bionlp-uima) is basically one step away from being a powerful, useful tool that the bioinformatics community could really benefit from. That step is really to do the beauty work - to make an application for people rather than just a code collection for hackers. The project reminds me a lot of my all time favorite open source software project WEKA - the Waikato Environment for Knowledge Analysis. WEKA contains implementations of thousands of machine learning algorithms along with a tools for experimenting with them. The key difference is that it has a stable click-and-run user interface to provide access to those tools (though you can still access, change, and learn from the large Java stack that runs it). If the BioNLP code was wrapped up in such a framework I suspect they would get many more users, I would certainly be much happier ;).

(Note, this my first foray into this kind of unstructured information extraction. If you know of better ways to do it, please do let me know!)

one last Connotea complaint

2010-08-17T09:41:00.001-07:00

Dear Connotea,

Why must you plague me with your well-designed, pleasingly styled human interface, your elegant RESTful API, your fascinating accumulation of knowledge and your GODAWFULLY UNRESPONSIVE SERVER!!!

What on Earth is wrong with you???

Guessing GOAs

2010-07-27T15:54:00.000-07:00

Guess what! I'm looking for a database of gene annotation guesses - the Not-Quite Gene Ontology Annotation database (NQ-GOA).

We have built a system for generating guesses of new gene ontology annotations for genes. I would like to see how our results compare with other systems' guesses because, my guess is that the more we are guessing the same thing using different guessing instruments, in general, the more we are likely to guess correctly, I guess.

So, where are your best guesses?

Google buys Metaweb

2010-07-16T14:38:00.000-07:00

If there was any previous doubt about whether 2010 will be known as the year that the Semantic Web came of age, this acquisition seals the deal.

I wonder if my life would be different if I had listened to my wife's advice and walked up to Danny Ayers at sci foo in 2008 and asked for a job... His three word intro "just, an, engineer" still inspires me.

Well done Danny and friends!

Probabilistic Partioning of the Gene Ontology - anyone?

2010-06-30T08:48:00.000-07:00

Yesterday, Andrew Su and I were wondering about disjointness in the Gene Ontology. We mused, "are there concepts in the GO that are mutually exclusive from the perspective of the gene products that they are associated with?" In other words, if a gene is anointed with a particular annotation, are there any other annotations that are thus rendered impossible for that gene?

With a small amount of searching, I found that the GO consortium is researching the addition of declarations of disjointness such as "intracellular is disjoint from extracellular". However, such statements about cellular space don't necessarily transfer to statements about the genes that occupy those spaces. This is clearly illustrated by 390 human proteins that are annotated as both 'intracellular' and 'extracellular' (and is stated explicitly in the GOC definition of 'Disjoint from'). So, even when/if that gets finished, its not going to help with our question. Furthermore, it seems that searching for the strict, perfect boundaries of formal logic might not be the right approach to this anyway.

What we would really like to see is a probability. Given that a gene is annotated with a particular concept from the GO, what is the probability that it could also be annotated with another GO concept? Concept pairs where the probability of co-annotation approaches zero would represent the (probably) disjoint biological classes that we were originally thinking about.

Naively, it seems this would be fairly easy to generate using the GO annotation database. Just build a contingency table for each pair of GO concepts (avoiding sub/superclasses) with counts for numbers of genes annotated with each concept and compute the statistics. If anyone would like to implement this (even better implement it in a non-naive way that accounted correctly for priors and chose the best test statistic) we would:

be very happy to use the results in the context of another investigation that is already in progress and that motivated the question...
be very happy to share authorship on any papers that might come out of the probabilistic partitioning service or the mysterious study already in progress...

If this work has already been accomplished and I've missed it, please let me know! If you are interested in working on it (or have already started working on it on your own) also please let me know!

-------------------------------
Partially related efforts but not quite what we are after...

Peter Mika at SD Semantic Web Meetup tomorrow

2010-06-18T07:57:00.000-07:00

For anyone in San Diego with an interest in the Semantic Web, tomorrow's meetup is a must. The presenter for the day will be Peter Mika from Yahoo! Research in Barcelona, stopping in on his way to SemTech. Peter gained early career fame for investigations into the relationships between social networks and the semantic web such as the award winning "Ontologies are us: a unified model of social networks and semantics". Abstract from the Meetup site below, hope to see you there.

"While current search techniques aim at ever more sophisticated methods for searching over hypertext, the Semantic Web promises to break boundaries in search by transforming the content itself into a form that is more easily processable by machines.

In this talk, we will discuss some of the possible technologies for annotating content for machine processing and showcase some of the ways that semantic annotations can improve the search experience for users. In the first part of the talk, we will describe existing and upcoming formats for embedding metadata inside web content, including microformats, RDFa, Facebook's Open Graph format and HTML5's microdata. We illustrate these formats with practical examples, show some of the tools that can help authoring and give guidance on best practices and point out some of the pitfalls, drawing on our experience from working with large publishers. In the second part of the talk, we will discuss research in semantic search and demonstrate its applications in Semantic Web search engines and web search engines with semantic extensions, including Yahoo's Search Monkey and Google's Rich Snippets."

Gene Wiki Jamboree 1 - post mortem

2010-06-07T08:14:00.000-07:00

Last week I had the honor of leading a Gene Wiki Jamboree at the annual meeting of the FaceBase Consortium (an association of researchers focused on understanding the genetics craniofacial development). The FaceBase group is creating a centralized resource for sharing their data and their knowledge with the hope of generally cross-fertilizing ideas and specifically enabling analyses that span the domains and datatypes of the very diverse members in the group. As part of this effort they want to compile and share their knowledge about genes that are relevant to craniofacial development. Rather than, or perhaps in addition to, the creation of their own wiki they are hoping to do this in the context of the Gene Wiki - hence my appearance at their meeting.

In general, the goal of a 'jamboree' in the context of biology is to bring a group of people together to rapidly assemble their knowledge on some specific subject. To my knowledge, the first usage of the term in that fashion was the Drosophila Genome Annotation Jamboree at Celera in 1999 (please let me know if there was an earlier use). In that case, more than 40 scientists gathered together for 2 weeks to define the boundaries of the fly genes and to classify their function. Not being a company like Celera nor having a whole, brand new genome to work on, the size of our group and our time allotment was much smaller. In total we only had about 8 people and 90 minutes...

Given those constraints, this jamboree was really more of a tutorial. It took about 45 minutes to get everyone in the room set up with an account and through their first 'edit this page' experience. So in the end, not a lot of content was generated. But, everyone did succeed in making an edit, seemed happy with their experience, and were clearly much more likely to edit something in the future than they were before the jamboree. So.. I think it was a success given the situation.

A couple notes on the process:

We set up a friendfeed group so that people could ask questions remotely and could track what others were doing (we added RSS feeds for the contributors edits). This was not used at all. None of the participants was a previous friendfeed user and they were also completely new to the Wikipedia experience. It seemed that the attention load of learning the Wiki way was more than enough to occupy them. Also, everyone was on laptops which limited the screen space for a flow-interface like friendfeed. I still like the idea of the ff-group as a live view on the progress and thoughts of the community, but for an in-person jamboree we will have to think of a different approach to achieve that kind of collective awareness. Perhaps a separate projector with the flow view..
Principle Investigators with their own very busy agendas really aren't the best participants in a jamboree. To be more effective, we need people with more time to focus on the task (graduate students please..)
We need more time! Doing a good job editing the text of a scientific review article isn't the easiest thing to do, more time would really help. This is especially the case when working in an editing environment where, if you are adding references, the text you are typing starts looking like this:

VEGFR-2 appears to mediate almost all of the known cellular responses to VEGF {{cite journal | author = Holmes K, Roberts OL, Thomas AM, Cross MJ. | title = Vascular endothelial growth factor receptor-2: structure, function, intracellular signalling and therapeutic inhibition. | journal = Cell Signal. | volume = 19 | issue = 10 | pages = 2003–2012 | year = 2007 | month = Oct | pmid = 17658244}}.

Yuck...

Watch out for edit conflicts. When some one is editing their very first page, the last thing that you want to happen is for some one else to be editing it at the same time. While it could be construed as a useful lesson, its really more than is needed for edit #1. This is something that anyone leading a tutorial like this should consciously avoid since its easy to accidentally create situations where it is likely - for example, by going to a sandbox page or other and having everyone edit it at the same time or even just posting a list of pages to edit (most people will start on the top).
Finally, it is vital to have some kind of central coordinating page where links to help resources and to-do lists can be assembled. I probably didn't give this enough thought on this run, but will be sure to make this a priority in my preparations if I get the chance to lead another Jamboree.

Gene Wiki Hairball 1

2010-05-27T13:15:00.000-07:00

Here is a view of the hyperlink network in the Gene Wiki.

Generated as follows:

Query Wikipedia for articles in the Protein category (using the API with some help from the Bliki java client library).
Extract links to other Wikipedia articles
Build link network in Cytoscape (organic layout)

Pretty I think, now what do we do with it?

New home with Gene Wiki

2010-05-20T14:28:00.000-07:00

Hello from my new home away from home at the Genomics Institute of the Novartis Research Foundation (henceforth known only as 'GNF')! I have come back into the warm, comfortable folds of mother Science as a postdoc in Andrew Su's group where I will be working on the Gene Wiki (no not WikiGenes, not WikiProteins, not WikiPathways and not any of the other biowiki wonders).

The broad purpose of the Gene Wiki effort is to describe the function of all human genes. While most other thrusts in this direction emphasize structure and control for capturing these annotations, the Gene Wiki project is guided by the undeniable fact that:

"Data without structure is still valuable, but structure without data is not"
(Andrew Su)

Based in part on this premise, the creators of the Gene Wiki made the choice to start their work directly in the context of the mother of all Wikis and one of the largest distinct sources of unstructured data on the Web - Wikipedia itself. This, of course, presents distinct advantages and disadvantages. On the plus side, there were already many users, pages, and most importantly editors before the Gene Wiki had a name, the Wikimedia foundation handles all of the infrastructure, and the wiki articles profit from ridiculously high amounts of Google Karma - more or less ensuring that the vital flow of people through the pages will continue. On the minus side, the lack of direct control over the technical infrastructure sharply limits the introduction of new interfaces for adding or interacting with content beyond the fairly tight constraints of WikiText.

Now, you may find it odd that some one like myself that has spent most of the past half decade trying to figure out how to add more structure to bioinformatics resources on the Web and complaining vociferously about the lack of interest in doing so (at least according to the latest Web standards) would now be working in the amoebic, almost structureless world of the wiki but fear not - the central aim of my work here is to figure out how to get more structured data out of the articles in the Gene Wiki...

Jerzy Lewak at SDSW Meetup

2010-04-06T21:50:00.000-07:00

Image via Wikipedia

I just got back from another interesting San Diego Semantic Web Meetup, here are my notes before I forget.

The presenter this evening was Jerzy Lewak, a former theoretical physicist, a professor emeritus at UCSD, and cofounder of (at least) Nisus Software and SpeedTrack Inc.. (Once again, I continue to be impressed at the level of the speakers that are recruited for these events!). Jerzy presented the history and current applications of a novel human interface for databases that he calls GIA (Guided Information Access) that is made possible by the underlying TIE (Technology for Information Engineering) framework.

Jerzy began his talk as a true computer scientist by defining 'The Problem' that originally inspired this work (back in 1991). At that time email was just starting to pick up steam and already the task of finding old emails was becoming unmanageable. So, he set out to find a better way. (Always nice to be working on a solving a problem for yourself.) His search for a solution took a decidedly familiar path:

Hmm, why don't I set up some nice hierarchies of concepts to place my emails into so it will be easier to find them later?
Darn.. that really isn't working out very well. The more data I get, the harder it is to organize and I keep running into the problem that almost every single item in my collection might be placed under more than one category. Perhaps a physical filing cabinet is actually a terrible thing to base a completely virtual information storage and retrieval system on... (Though he didn't bring it up, he was describing exactly the same thing that Clay Shirky got so excited about in the 'Ontology is Overrated' essay in 2005 that - in turn - got me all excited in 2006, except of course Jerzy was thinking in 1991. And, as pointed out to me by my LIS friend Joe, this basic problem and the following conclusion were pretty well fleshed out by Ranganathan in the 1930's...).
After experimenting with plain content search (a la Google) he arrived at faceted classification as the most powerful and flexible way to describe and access data.

So far so good, I am paying attention.

Now the problem arises that many combinations of facets actually produce zero results. As he noted, just 200 facets (he uses the word 'selectors') is enough to uniquely describe every particle in the universe. This became the real problem. The breakthrough that got him going and has led to SpeedTrack and everything else he presented was the idea of dynamically limiting the potential facets based on those that are already selected. In the interfaces he demonstrated, he would:

Choose some database field like 'last name' and type in a name like 'Smith'
Show two things - one, the number of results went down and two, the number of possible values for the other fields (e.g. 'first name', 'height', 'date', etc.) would immediately be constrained to only show values for objects linked to Smiths in the database.

His interfaces might be described as an advanced multi-parameter type-ahead. They work by guiding the user (GIA) in the creation of potentially very complex queries that are guaranteed to return results. This is achieved by dynamically exposing the underlying indexes that drive boolean queries.

The system works on both structured and unstructured data (he showed a quick example of newspaper articles) but requires fairly heavy manual labor to get it started. Overall it looked like it would be useful and fun to use and I can imagine many potential directions they could take it.

My only complaint for the talk was that there was absolutely zero Web in it - no mention of any native ability to consume or produce RDF, no mention of OWL, no discussion of scaling possibilities, and the proverbial elephant in the room of large-scale data access was left more or less untouched. I guess that might be one sign of a good talk - I got interested and it left me thirsting for more..

note to self

2010-03-14T21:15:00.000-07:00

Someday attend South by Southwest... It seems inspirational. And fun.

bio2rdf 2 biogps

2010-03-10T21:46:00.000-08:00

Last Friday I had the pleasure of having lunch with with Andrew Su of the Genomics Institute of the Novartis Foundation. Among other things, he introduced me to one of his projects called BioGPS. BioGPS is an interesting, kind of minimalist approach to gene-based data integration. Essentially, it allows you to register gene-related 'plugins' that other users can assemble like an iGoogle home page. Each plugin amounts to an html-producing url that contains one of a variety of gene ids as a parameter. So you might have a plugin for ncbi gene, another for kegg, etc. and they are all displayed together using a very smooth, interactive iframe canvas.

This is clearly useful to many people (they get about 150,000 pageviews/month) but its flexibility is limited by the way it currently accesses information - simply by gathering HTML from existing web pages. Since many pages have overlapping content there is inevitably (screen)wasteful duplication in the aggregate view. As others have said before, a little bit of semantic web could go a long way to improving this resource - and, because of the way the system is built and the way SPARQL endpoints work, its very easy to do it.

So, the idea is that you could take a sparql endpoint (that yields html as an option), write a query with a gene as a parameter, capture the url that contains the query and then you have a very specific kind of plugin that only shows precisely what information you want. By assembling a collection of these you could produce a view on the gene information space that was very precisely tailored to individual needs.

I made a simple example of this pattern with the plugin "OMIM disorders where gene is linked to pathogenesis" which you can see in their plugin library.

It hits this endpoint
http://atlas.bio2rdf.org/sparql

with this query
PREFIX omim: PREFIX rdfs: select distinct ?OMIM_disorder where { ?s omim:PATHOGENESIS ?o . ?o bif:contains "VEGF" . ?s rdf:type omim:GeneticDisorder . ?s rdfs:label ?OMIM_disorder }

where the VEGF would be replaced by the gene that you were researching.

Here is an example BioGPS view composed of four plugins. My bio2rdf-sparql example is there on the top right.

To really do this properly, I think you would want to build a little helper application that would help users compose the queries and would allow for some basic formatting options for presenting the results of these SPARQLing BioGPS plugins.

What would you do with a semantic pubmed?

2010-03-09T19:44:00.000-08:00

If all of the concepts buried in the text in PubMed were dug out, stitched together into a semantic network and placed into a system that could rapidly provide answers to database-style (e.g. sparql) queries over that graph what would you do with it?

I was recently posed this question and have been scratching my head to come up with a very good answer. I can't seem to get past the first thought that popped into my head. My initial idea was to build an interface that would allow any user to gain answers to this fundamental question

"how is concept A related to concept B".

As I saw demonstrated a few weeks ago, the technology exists to execute such queries but I personally have never seen a web interface that would allow untrained end users to explore very large knowledge bases in that fashion. I imagine something as simple as two autocomplete boxes for the input and then a variety of presentation styles for the output (graphs, text, etc.).

Any other ideas ? I might get a chance to build something like this soon.

-Update (link to small related friendfeed thread)

Dreamers of Day - repost

2010-02-22T17:42:00.000-08:00

I came across this quote in a Bokardo post and it stung. Here it is as a reminder for myself not to take too much pride in unrealized visions; regardless of how clever they might make you feel at the time you have them they are ultimately nothing until they are brought to life.

“All men dream; but not equally.

Those who dream by night in the dusty

recesses of their minds

Awake to find that it was vanity;

But the dreamers of day are dangerous men.

That they may act their dreams with open

eyes to make it possible.”

T. E. Lawrence

Blue Tweets

2010-02-05T19:26:00.000-08:00

I find it interesting and slightly annoying that Twitter seems to have replaced social bookmarking services. Perhaps its just within the admittedly small community that I follow on occasion, but it seems to me that the main reason for a very large percentage of tweets is simply to share a link with a brief comment. While putatively for personal information management, a major reason that people use social bookmarking services (like Del.icio.us , Connotea and so forth) is to share (hence the 'social') [1] and I frankly think they do a better job of it then the ambiguous tinyurls and bit.ly's taking up valuable character space in tweets.

But... Connotea and its brethren are the past and Twitter is the now so here is a suggestion for making the twitter of tomorrow a tiny twit better. Perhaps it should be possible to simply make the entire text of a tweet a link. Same function, more visually appealing interface and more characters left at your disposal.

Just a thought.

[1] In my own experience sharing has really become the primary function, with refinding old bookmarks almost always easier with a new Google search then via my own tags (even my fancy tags).

Jans Aasman From Franz Inc. Presents at San Diego Semantic Web Meetup

2010-01-31T12:46:00.000-08:00

Yesterday, about 20 people gave up 4 hours of sunny Saturday afternoon to gather in Carlsbad to listen to Jans Aasman present a tutorial on AllegroGraph - what Franz Inc. is touting as "Web 3.0's Database". The attendees, some of whom drove all the way down from Santa Barbara, included a number of people interested in semantic search and search engine optimization. There were also software developers, CEOs of software development companies, bioinformaticians, an orthopedic surgeon, business development specialists, and entrepeneurs. A challenging crowd to give a tutorial too that all would find useful.

Since I ended up having to take my notes on my iphone and the presentation lasted nearly 4 hours, I'm only going to hit a couple of the things that stood out for me here. For more info. on the products presented, the Franz Inc. website has extensive information - and free downloads! I also hope to find a link to the slides at some point.

First of all, wow - well done organizer! Its not everyday that you get to hear a presentation by the CEO of one of the oldest (is it the oldest?) living companies involved in artificial intelligence. (Yeah I said it.. thats what we used to call this stuff.) Franz has apparently weathered the storm and now that semantics is coming back to life under a new name - has come out doing very well with clients including: Lilly, GlaxoSmithKline, Adobe, Raytheon, Kodak, Boeing, Cisco, Mayo clinic, Novartis, and many others. These clients are paying for the pro versions of tools like AllegroGraph and Gruff as well as consulting services related to the use of these tools to solve specific problems. Jans began his presentation with a really rapid explanation of why these companies are spending their money like this rather than on continued development with the encumbent technology - relational databases.

Jans suggested that triple stores like AllegroGraph are more useful then relational databases when any of the following constraints are met:

You have many classes of complex objects
The properties/definitions of these classes change frequently
You want to work with rules/reasoning
You have a big graph in your data that you want to analyze

Of those he hit on #2 the most often. Basically, if you can perfectly conceive of the entire world that you need to model at the time you create your database schema you are probably going to be better off using the more mature rdbms technology. If not - and this is the likely case for more projects then your database architect would have you believe - the triple tech. will make change much easier.

After that brief, high-level motivator he dived into a fairly extensive series of demos/examples. While the subject matter varied from news stories to clinical trials to digital photo websites, there were two consistent themes that I found interesting.

Every example he gave involved a named entity extraction step at one point or another. This indicates that, while there is a growing amount of structured data out there, there is still way too much information held in text to ignore when doing any real analysis. So.. you are going to need to get friendly with some people that do natural language processing - perhaps like this company Alchemy.
Most of the demos were conducted in Gruff, the visual interface to AllegroGraph. In Gruff, he repeatedly showed off the very cool trick of selecting two, apparently unrelated nodes in the graph (say a two different drugs) and asking for connections between them. AllegroGraph responded to this request ~instantaneously with relevant and useful results. Very impressive technically (this was a graph with millions of triples and the system was running on his laptop) and I think very useful for discovery.

Wish I could write more but its the weekend after all.. For more information on this stuff, go to Franz inc., grab yourself a free download or two, and try it out yourself! I'd be curious to hear how you find it.

Eric Little presents at San Diego Semantic Web Meetup

2010-01-19T22:08:00.000-08:00

Just returned from my first Semantic Web Meetup in San Diego, here are some impressions:

In comparison to the Vancouver SW meetups that I have attended

Quite a different crowd. Larger group, zero t-shirts, generally older,
stronger industry representation, heavily weighted towards biomedical
folks - even an actual surgeon in attendance.
As with everything in Southern California, it involved a fairly long drive.
If I was a small fish in Vancouver, I'm a microbe here.

A few bullets from the (very good) talk

The presenter, Eric Little, did his PhD with Barry Smith (in formal biomedical ontology) and came from a background in philosophy. He is now the chief knowledge engineer at CTG - a fairly large (revenue in the 100's of millions) Information Technology and Consulting company.
He talked for a while about how ontologies should provide "actionable intelligence" and gave some nice stories about the benefits gained when links are formed across semantic groups - for example, they created a diagnosis ontology by linking the disease ontology to the LOINC (logical observations, identifiers, names, and codes). Kind of the same old story here, but good examples.
He then got to the meat of the talk (according to the abstract), which was about their MedMap application. He, correctly IMHO, pointed out that pretty much every attempt at an interface for dealing with large ontologies more or less sucks. No offense... He then proceeded to claim that their solution in the MedMap application was much better. The comparison to things like Protege and Concept Map builders was a bit unfair as the MedMap application is targeted towards end users rather than ontology engineers but, I can't say I disagree with the picture he painted.
The MedMap application is a very nifty widget-based approach built using Adobe Flex on the top and with combination of Top Quadrant products underneath (notably SPARQLMotion). It looks a lot like the iGoogle interface, with a user-customizable collection of what he called "analysis portlets". Each of the portlets has a dedicated task, but they all sit on top of the same massive knowledge base. The examples he gave were clinical - for example, he showed one scenario where patients were being automatically classified into different risk groups with a nice little graphical representation of the different classes; however, the technology and many of the interface modules are clearly fairly general purpose. He mentioned that they are in fact working in quite different domains such as the petroleum industry.
To get your own MedMap-like solution for your problem, starting rates are on the order of $300-400,000 for about a 3 month long push for a first usable product. They are after big fish like the health insurance industry.
His opinion was that the only companies that are likely to be interested in buying semantic IT solutions are those that a) have a lot of money, b) are more forward-thinking than average, and c) have already tried other approaches (e.g. relational databases) and watched them fail.

Personal notes/reactions

The sheer slickness of that Flex interface still has me all in a quiver.. It makes me want to delete everything we have been doing on my current attempt at a semantic application (I am now embarrassed to put in the link) and start over - or perhaps it is just time to concede that its impossible to build Ferraris like I what just saw without employing a staff of people that know what they are doing.
It was cool to reconnect with Barbara Starr, my former supervisor at SAIC - which was prior to my MSc, my foray into bioinformatics, and my PhD! Grey hairs...
It was also very cool that she is about to start working on an application of the GoodRelations ontology by Martin Hepp. I am proud to say that Martin sent me unsolicited positive feedback about my first research project in grad school several years ago. He concurred with my claim that ontology development/maintenance practices were too expensive given typical knowledge engineering practices and that techniques that tapped into the collective knowledge of the web (a la 2.0) were the way to go. (Note that this contention is in stark contrast to the ontologies of "reality" from Barry Smith.)

All in all a very good meeting and I am very much looking forward to the next one.

Selling Market Research Slimily

2010-01-16T17:12:00.000-08:00

One of my current endeavors has drawn me into the bizarre world of "corporate intelligence". This basically means information about other companies that you can use to understand your competition and find partnerships. As I seek to understand this strange and dangerous new land, I've been spending a fair amount of time exploring websites like http://www.biomedtracker.com/ and within the confines of such websites I occasionally see tantalizing links like this:

These wonderful images get me all excited - "I will finally learn of the magical secrets held beneath! " - but then they invariably lead to something like this:

But I don't want to contact Bobbi Sue Richardson! I just want to learn what your product actually is! Why must you torture me so? Why can't you be more like the folks who are doing so well at 37signals who "gasp!" allow me to view their product demo videos for things like basecamp with the click of a button.

While these walls frustrate me, they also fascinate me. Why all the cloak and daggers? There must be a reason.. Any ideas?

2009 in haiku

2009-12-31T18:21:00.000-08:00

2009 was a big year for me. Aside from the minor details of finishing my PhD, having my first child, and moving back to California, I started to use my 2 year old Twitter account! Yes, I know, I'm so unique. Anyway, when I started tweeting I had the grand idea that I would make all my tweets Haikus. This didn't last long but I still occasionally try to slip one into the stream. Inspired by my Twitter inspired poetweetry, I present my year 2009 in Haiku.

January came

with it furious writing

then deliverance

Snowy mountain trip

powdery time of waiting

good news  

Spring offers a test

and with it anxiety

happy its over

The city cooked

in underwear  I coded

while sweating light beer

Clouds return to me

a belly swells with new life

anticipation

Stork flies on cold wind 

screams of pain bring fear and joy 

intense genesis 

Riding on the storm

homecoming is imminent

California

Happy New Year!

Heading to China

2009-12-03T15:29:00.000-08:00

Image via Wikipedia

Tomorrow I am departing to attend the Asian Semantic Web Conference in Shanghai, China. I'll be manning the demo for my former labmate's project CardioSHARE (which I played no real part in building).

Looking forward to getting caught up on the latest from the semantic web research community. Out of the accepted papers, I am most looking forward to hearing "Merging and Ranking answers in the Semantic Web: The Wisdom of Crowds".

Hope to see you on the other side of the Great Firewall.

Birth and Rebirth

2009-11-18T17:09:00.000-08:00

Its now been 4 weeks, one hour and about 51 minutes since my son was born. Its also been about 6 months since I finished my PhD and about 6 1/2 years since I had a 'real' job. Perhaps its time to get on with things and sort out what I'm really going to be doing with myself. Here's an update on what I've been up to, opinions on my next steps would be welcome...

Having grown somewhat disillusioned with the academic world, I've spent the time since graduation working completely outside of it. Working with my father and his partner I learned how to use the Google App Engine while building a website/database for their company TrueIDapps. I've also been involved with the development of a semantic web based startup that is now going by the name FreeForm Information. Sadly I have not done any work in the domain of bioinformatics for a long time now.

While my post-grad projects have provided some knowledge of fun new techniques (I love the Cloud like everyone else now) and I am beginning to get a glimmer of an understanding of the process of starting and running a business, I have as yet to make a single penny on either project. Now, faced with the screaming, squirming reality of responsibility I am feeling the pressure to make some decisions about how I should be spending my - now much more limited - work time.

Do I :

Continue to try to keep myself involved with both FreeForm and TrueIDapps - in the hope that one of them will eventually pan out?
Focus my attention on TrueIDapps because I have two other fulltime coworkers versus one part-time partner, because it always feels good to support the family business and because it seems that it is close to making money?
Focus my attention on FreeForm because I find the project more interesting, more relevant to my past experience and more inline with the work I would be trying to do if I managed to find a real job?
Drop both projects and figure out a path towards a real* job?
Commit to being Dr. Daddy, buy some baby formula and get Dr. Mommy back to work ?

I have the benefit of my wife's savings and great family support so I could certainly last for a while before I hit the financial danger point but.. the pressure is mounting.

Your insights are most welcome!

*real job = a job with a salary, paid vacation, an office, - and a boss who tells you what to do.

mass perception

2009-10-12T14:05:00.000-07:00

If you've been to a concert or other live performance in the last few years, you have probably noticed a phenomenon like the one in the photo to the right. As the show gets started, a glowing school of digital cameras emerges out of the night like a swarm of fireflies and persists until the lights come back on. (Not quite as pretty as lighters, but at least they don't burn your thumbs.) When I see this I wonder:

What does the show look like from that guy's camera over there? I wish I could tune in and see, now.
What kind of creation will emerge when it becomes possible for artists to access all of those different electronic eyes and ears at the same time?

When live, phone-to-phone-to-Web video streaming becomes widely used, I think we will see some very new takes on live performance and I'm really looking forward to it.

---
Notes:

I took the photo at a fireworks show in the Butchart Gardens near Victoria, British Columbia.
If you are interested in live streaming technology that can work from phone to phone, check out a friend's company ZygoDigital.
If you want to read some extraordinarily prescient stories involving near future technologies like this - the ubiquitous internet connectivity aspect in particular - see Vernor Vinge's "Rainbows End".
Thanks to the comment on my last post, I am trying out Zemanta on this one.

Semantic media retrieval service? please?

2009-09-27T22:55:00.000-07:00

Here is an application of semantic Web technologies that I would like to have. Please make it for me so that I don't have to. When I finish writing this post,

I would like to press a button that said "Enhance?".
When I pressed the button, the application would read through the text and identify terms, phrases, or other conceptual nuggets that it 'understood'.
These concept nuggets would then be used to find stock / open access images (and videos, etc.)
Where a likely candidate set of images was identified, they would be displayed such that I could quickly choose which, if any, that I liked
When I agreed to keep one, it would be embedded in a reasonable location in the text and I would very rapidly go on with my life, but with the added joy of having authored a much more entertaining piece of online personal history.

This thought crept into my mind after reading through Joey de Villa's post about joining Microsoft which is shot full with entertaining media enhancements to the text - which likely took a non-insignificant amount of time for him or his team of personal assistants to put together.

Pictures are indeed worth many words, but how many $$$'s? Perhaps you might even be able to make money with such an app by using it to sneakily sell professional photos and other content.

While you are at it, could you please provide the same text-to-media service in a non-embedded application so that when I needed a clever portrayal of a concept like 'failure', 'success', or 'mass collaboration', for a presentation I could quickly look one up. I might even be willing to by it if the content was good and the price was reasonable -> in a world where I could almost certainly find what I needed by spending a little more of my own valuable time looking for it.

bottle rockets and businesses

2009-09-15T21:32:00.000-07:00

Last night I attended a Demo Ignite Camp hosted by Microsoft as part of their techdays conference, organized by Boris Mann of Bootup Labs, and headlined by Joey de Villa (Developer Evangelist at Microsoft in Toronto).

Joey initiated the event with the following video which I insist that you watch immediately.

For the (non-obvious) explanations of why this video was the perfect start to the session, see his article about joining Microsoft and the blog post about the video that inspired him to do so.

Butt Rocket - Watch more Funny Videos

Strange reality of academic workspace

2009-09-14T22:51:00.000-07:00

My sister forward this job posting to me.

We are looking for a researcher or consultant to participate in the
collaborative development of Neural ElectroMagnetic Ontologies (NEMO). This
NIH-sponsored position would involve developing ontologies for
representation of patterns in event-related brain potentials (ERP) data
that reflect various aspects of language processing.

The Associate would work with members of our International NEMO Consortium
(experts in EEG and MEG studies of language) to develop, manage, and curate
ontology and database structures for this project. Please see the website
for more information, and contact Gwen Frishkoff if you have any questions.

The ideal candidate would have a background in cognitive neuroscience,
experience with database design and curation, and/or familiarity with
Protege/OWL ontology development software...

Sounds like an interesting job that matches my background pretty well, but read on..

Salary will be commensurate with experience in the range of $30-$40K per year.

Perhaps I'm just greedy...

Freebase authors semantic web book

2009-08-02T11:39:00.000-07:00

I just noticed this post on the freebase dev blog - announcing the release of a new book about "programming the semantic web" written by freebasers Toby Segaran, Colin Evans, and Jamie Taylor. I haven't picked it up yet, but based on the description there and in the amazon reviews, it looks like a(nother) nice explanation of the concepts involved in the semantic web as well as a set of practical programming examples based on the W3C standards (OWL, RDF, SPARQL, etc.).

I found it surprising that freebase was not mentioned anywhere in the brief description given in the post or on Amazon.

publisher removal

2009-06-05T09:36:00.000-07:00

In my last post, I mentioned offhand that I could not access a PDF (about an ontology for autonomic license management) without paying a $29 fee to Springer. Though the post was not a direct request for help running around this paywall, I have now received the pdf from 5 different people - several of whom I have never met before.

Clearly, the (micro) community that read that post believe that research articles should be shared in an open-access fashion and that it is both wrong for publishers to charge access fees and right to sabotage the publishers via peer-to-peer exchange of such articles.

I'm wondering if this micro-community (that is you) would feel any differently if the fees paid for such articles went directly to the researchers that produced them rather than to an apparently irrelevant publisher ?

(Also, I wonder about the Radiohead style "tip jar" approach. This would allow you to read the article first and then contribute a payment afterwords if you felt that the research in the article was worthy of supporting.)

licenses and linked data

2009-06-02T12:39:00.000-07:00

One of the major downsides of working outside of academia is that I now have to pay much more attention to licenses. No longer can I just grab whatever data I like, do something fun with it, and try to publish what I did and move on. Now I need to know - very specifically - what I am allowed to do with what so I can reduce the possibility of being sued and so that I can put up appropriate "powered by bla bla" messages. Not being fond of reading legal agreements, this is a drag.

So I thought to myself, "there should be a legal ontology for linked data!". That way I could tell my data harvesting program to ignore (or hide I suppose) any data that I wasn't legally allowed to put into my new for-profit-maybe-someday-I-hope mashup and thus never have to bother reading those agreements again.

I am not the first to think of this of course. Here is a nice-looking abstract attached to a paper that I would like to read but am no longer allowed to read for free.

"The license agreement can be seen as the knowledge source for a license management system. As such, it may be referenced by the system each time a new process is initiated. To facilitate access, a machine readable representation of the license agreement is highly desirable, but at the same time we do not want to sacrifice too much readability of such agreements by human beings. Creating an ontology as a formal knowledge representation of licensing not only meets the representation requirements, but also offers improvements to knowledge reusability owing to the inherent sharing nature of such representations. Furthermore, the XML-based ontology languages such as OWL (Web Ontology Language) can be user friendly for the non-developers who are often those responsible for implementing and managing such license agreements. This paper shows our use of ontology to represent the license agreement in a development prototype. The ultimate goal is to build ontology for the license management domain that will facilitate autonomic knowledge management. Knowledge based on such ontology can then be shared and utilized by many types of license management system. "

What do you think? Is it worth it to pay Springer $29 for the 1592kb in that paper?

CWA at the YMCA

2009-05-11T10:19:00.000-07:00

Somewhere high in the air between New York and Minneapolis, my first stop on my way home, I feel compelled to explain a few things to myself. Why on Earth have I just spent the last several nights living in the YMCA in Flushing, New York? Why did I decide to go on my first self-funded professional excursion at a time when I have no income and very little savings? What did I hope to get and what did the trip deliver?

The inspiration for this minor adventure was the inaugural meeting of the Concept Web Alliance (CWA) at the New York Hall of Science. The mission statement of the CWA (written partly at this meeting) is as follows:

“To enable an open, collaborative environment to jointly address the challenges associated with high volume scholarly and professional data production, storage, interoperability, and analyses for knowledge discovery”

The idea is to form an alliance of like-minded researchers and science publishers interested in sharing knowledge in a computationally accessible fashion (i.e. not plain text and such that information from multiple sources can easily be integrated and interacted with). The basic building block envisioned for these efforts is the ‘triple’ – a Concept-Relation-Concept structure. (The word ‘triple’ and the interesting new verb ‘triplification’ - meaning to convert some non-triple-structure like text into a set of triples - were almost certainly the most commonly uttered words in the presentations at the meeting.)

For those familiar with semantic Web standards such as RDF (a generic triple-based language for representing and sharing information) and OWL (a set of languages for representing knowledge in the form of ontologies) it is perhaps most interesting to consider what is not present in a CWA triple and what was not discussed at all in the public portions of the meeting. The following words never came up ‘description logic’, ‘axiom’, ‘class’, ‘reality’.

The intended materialization of the Concept Web vision - at the moment - thus seems to be an open collection of informal (non logic-based) concept representations, identified by URIs retrievable on the Web, that can be linked together to form semantic networks. This graph-structure could be queried (e.g. using SPARQL) for the ‘facts’ that it would contain where each such fact would be linked to extensive information about where it came from (who (or what algorithm) suggested it, when, and with what confidence). Interestingly, this is very similar in its flexibility, lack of built-in reasoning, and its strong notion of provenance tracking to the Freebase model.

While some of you who like to work with reasoners and OWL or who think that it is better to talk about ‘universals’ and ‘particulars’ than it is to talk about ‘concepts’ may find this lack of formality a little disappointing, I am growing more and more enthusiastic about it because fits the publish-then-filter nature of the Web perfectly. We see again and again that once information is out there on the Web, its value increases tremendously. (In fact, many very smart people seem to think that when there is enough text and other unstructured data online that is all we will really need to solve most of our information problems.) By providing a very low barrier for entry and then focusing computer science efforts on handling the noise, complexity, and the conflicts that will inevitably arise (the filter part), I think this triple-publishing approach has great potential to push research in a productive direction. In particular, I think it will push people to spend more time working on other, more flexible modes of inference that don’t just die when a logical conflict is detected - all that squishy probability stuff that the semantic Web has managed to ignore for so long and that happens to be the stuff that makes almost all interesting AI-like technology work now. Furthermore, the triple-focus absolutely does not stop groups that participate in the CWA from making use of approaches grounded in formal logics in their own development.

While there may or may not be good reasons to take one philosophical stance over another when creating knowledge bases, the fact will always remain that there will be conflicts of opinion about this. When dealing with a small group, it may be possible to convince or force acceptance of a particular world view, but it is not IMHO going to be possible to enforce something as arguable as the philosophy of the representation of the nature of being on the scale of the Web. By focusing on the smallest possible units, the triples, and leaving the more precise formalizations and the philosophy out of the vision as much as possible, the CWA might make it possible for a diverse, interoperable ecology of knowledge bases to emerge and co-exist. Ideally, those who wish to make use of, for example - description logic reasoning, should be able to benefit from the pool of URIs in the Concept Web if for nothing other than for the many multi-lingual labels and textual definitions that will be associated with each of them.

It is still very early days for the CWA – probably far too early to really speculate too far about the consequences of its basic technological approach as even this approach is still very much up for debate. Still.. I’m not sure exactly how to express this, but the meeting smelled good. There were enough capable, powerful, enthusiastic people together in that room that seemed to have enough of a shared vision that I think it is very likely that something of that vision is likely to come to life.

So, was it worth it? I think that it was in the end. I got an early look at something that might provide solutions to many of the problems that I’ve spent the past several years of my life thinking about (the social construction of a biosemantic web). I got to reconnect with old friends and make some new ones. I had a chance to see New York for the first time (the scale of which blew my mind). And, last but not least, it just might be the last such academic event I get to take part in. Depending on the choices I make and the dictations of the wheels of fate I may be in the process of losing the privilege of working in the ivory tower. If it was indeed my goodbye to the community of scholars, it was a good one.

So yes, it was a worthwhile trip and it remains an exciting time to be thinking about the Web - concept or otherwise. (and the YMCA wasn’t really so bad in the end ;).

You can follow - and perhaps influence - the evolution of the Concept Web on their blog.

CWA live webcast today

2009-05-08T08:03:00.000-07:00

In case you want to follow the Concept Web Alliance meeting happening right now, it is being webcast.

Annotation evaluation

2009-05-06T09:15:00.000-07:00

A few years ago, a friend of mine was an author on an article about global gene coexpression analysis. In the article, they looked across many experiments to find pairs of genes that consistently shared patterns of expression (indicating likely co-regulation). As one of the validations of their globally co-expressed gene-pairs, they measured whether or not the sets were also co-annotated with the same biological process label from the gene ontology. The idea being that if they were co-expressed they should be more likely to belong to the same biological process.

This suggested to me at the time that this kind of analysis might somehow be turned on its head to provide evidence regarding the quality of the annotations themselves. Basically each annotation would be treated as a hypothesis and then continually subjected to testing. So the assertion GeneA is part of Biological Process Y would be strengthened if GeneB was co-expressed with GeneA and was also a part of the same biological process. I imagine there would be many other forms of such evidence.

Just wanted to jot that down..

The big apple

2009-04-24T09:36:00.000-07:00

After much hemming and hawing I've decided to make my own way to New York for the inaugural meeting of the Concept Web Alliance. It will be my first conference outside the $helpful$ cocoon of academia and my first visit to the home of the statue of liberty. I'll try to write more about what the CWA is after I find out more at the meeting.

My travel plans have me free for meetings - business or tourist - in NYC on May 7 and 9 (with the key meeting day on the 8th). If any readers are going to be there, let me know!

bad day programming

2009-04-23T07:27:00.000-07:00

You know you've had a bad day programming when, at the end of the day, you find yourself manually editing the text of a file that looks like this...

The amazing thing is that the edit worked!

For potential fellow sufferers:

On a Mac, the urllib2.py Python module guesses about your proxy configuration by looking at the file here on the right called com.apple.internetconfig.plist (in /User/you/Library/Preference/)
If you are having a problem creating connections with urllib2 , try having a look at what proxies it thinks exist like this:

>>> import urllib2
>>> urllib2.getproxies()

If it comes back with something unexpected like 'http://evil.proxy.bug:8080' then you will need to get rid of that offending string in the com.apple.internetconfig.plist file. The file will open and look a little prettier in the plist editor; however the editor has no search function and I couldn't find the string - which I knew to be there - anywhere. In the end I just opened it in Emacs, found it, zapped it, and prayed that I hadn't hurt anything in the process. Sometimes that works...

Dissertation now online

2009-04-15T14:07:00.000-07:00

The full text of my dissertation, "Strategies for amassing, characterizing, and applying third-party metadata in bioinformatics", is now available via UBC's information repository. It is "manuscript-based" so each of the chapters except the introduction and the conclusion can be read and understood independently. (So there is really no reason for anyone ever to try to read the whole thing in its entirety.)

Here is the abstract:

Bioinformatics resources on the Web are proliferating rapidly. For biomedical researchers, the vital data they contain is often difficult to locate and to integrate. The semantic Web initiative is an emerging collection of standards for sharing and integrating distributed information resources via the World Wide Web. In particular, these standards define languages for the provision of the metadata that facilitates both discovery and integration of distributed resources. This metadata takes the form of ontologies used to annotate information resources on the Web. Bioinformatics researchers are now considering how to apply these standards to enable a new generation of applications that will provide more effective ways to make use of increasingly diverse and distributed biological information. While the basic standards appear ready, the path to achieving the potential they entail is muddy. How are we to create all of the needed ontologies? How are we to use them to annotate increasingly large bodies of information? How are we to judge the quality of these ontologies and these proliferating annotations? As new metadata generating systems emerge on the Web, how are we to compare these to previous systems? The research conducted for this dissertation seeks new answers to these questions. Specifically, it investigates strategies for amassing, characterizing, and applying metadata (the substance of the semantic Web) in the context of bioinformatics. The strategies for amassing metadata orient around the design of systems that motivate and guide the actions of many individual, third-party contributors in the formation of collective metadata resources. The strategies for characterizing metadata focus on the derivation of fully automated protocols for evaluating and comparing ontologies and related metadata structures. New applications demonstrate how distributed information sources can be dynamically integrated to facilitate both information visualization and analysis. Brought together, these different lines of research converge towards the genesis of systems that will allow the biomedical research community to both create and maintain a semantic Web for the life sciences and to make use of the new capabilities for knowledge sharing and discovery that it will enable.

About the title.. I ended up using the generic "metadata" rather than something more specific because I needed a way to concisely capture things that range from Del.icio.us tags to classes in the Foundational Model of Anatomy. "Metadata" seemed to do the job, but it remains a little vague and therefore unsatisfying. Similarly, "third-party metadata" is broader than necessary. I don't touch institutionally generated third-party metadata - just what I guess you would probably call "socially generated metadata".

Enjoy.

shiny new pen

2009-04-14T15:11:00.000-07:00

Another page of conclusions, about 17 corrected typos, and about 7 more forms later and I have reached the penultimate experience of academic student life: a free pen from the university! After that comes the funny hat and then thats all she wrote.

For those silly people that were asking - assuming I did manage to find and sign all the needed forms - my dissertation should be available online within a few days.

Now... hmm

So long 23rd grade!

2009-04-06T09:16:00.000-07:00

After all the stress and the worry of the last few months I finally made it to the other side. Now, I suppose its ok to call me Dr. ~~Feel~~ Good if you insist ;)

I will refrain from doling out advice to my formerly fellow 23rd graders and just give a couple highlights for posterity:

Best advice I received before the defense: "Have fun with it" - Nicole Quenneville.

Worst beginning to an answer I gave: "That is a really hard question!" (laughter and heckling emerging immediately from the committee and the audience) - (Note that I did subsequently answer the question).

Questions I had the hardest time answering:

"What is bioinformatics?" - Francis Ouellette
"What emerging technologies do you think might have the biggest impact on the future of your research?" - Wyeth Wasserman

Best response from an examiner when I said I was still deciding about my next career steps (hint hint), "No way, I'm crowdsourcing everything from now on!" - Paul Pavlidis

Thanks one more time to everyone that has helped and encouraged me along the way - you all rock!

holy mashup, mashup man!

2009-03-18T14:49:00.000-07:00

I was recently accused by an associate of being a "mashup master" (because I never actually make anything from scratch I guess). I don't think I really fit the title, but this Kutiman fellow absolutely does. The music video below is entirely produced from the integration of different utube videos. I'm no expert in music or in music-mashups, but I think this is pretty special.

This shou serve as an inspiration for anyone who is thinking about what they might do with music or other forms of data on the Web. (Listen for at least about 45 seconds - the cool thing is how the song develops.)

tightening the knowledge cycle

2009-03-17T10:04:00.000-07:00

I've recently returned from a trip to the Leiden University Medical Center where I met with Barend Mons, Marco Roos, and Eric Schultes regarding the formation of what Barend is calling the Concept Web Alliance (CWA) (as well as some open positions within his groups at the LUMC and the University of Rotterdam).

In preparing for my presentation I read the WikiProteins article again (from Barend's group) and was struck with the similarity of the pattern I saw there with some of my old work and things that are emerging right now at Freebase and elsewhere. It seems that there is a very clear uptake of the following basic cycle in the context of building large knowledge bases on the Web.

Datamining (from text or other sources) generates many many candidate assertions
The assertions are presented to (many) people who manually correct/refine them

These two basic steps are being united everywhere right now. In WikiProteins we see natural language processing techniques seeding a semantic protein Wiki where people (presumably scientists) can correct/extend the predictions. In Google search we see the familiar products of datamining in the ranked result lists but these are now coupled with interfaces that allow users to 'vote-up' specific results. Freebase, which makes extensive use of automated knowledge acquisition techniques, is now unveiling a series of 'games with a purpose' like TypeWriter and Genderizer that are being used to validate the predictions of their algorithms (see a nice post about the Freebase process).

A somewhat more hidden indicator of this trend is that Dolores Labs - a company that seems to be entirely devoted to the application of Amazon's Mechanical Turk - is now hiring. Its hard to tell what they are doing internally, but it would be surprising if the interplay between datamining and the wisdom of crowds didn't play a strong role.

These trends will place increasing pressure on the scientific/tech community to come to a better understanding of the processes involved in motivating, coordinating, and aggregating the knowledge of many millions of minds in the formation of structured knowledge.

Authority, ontology, and Freebase

2009-02-23T11:05:00.000-08:00

Something I knew about but hadn't thought through until Kirrily Robert pointed it out to me yesterday at the Vancouver Freebase meetup is the importance of the 'as_of_time' parameter in freebase. This allows developers to access views of freebase data at any point in time. The important consequence of this is that it makes it easy to create authoritative versions of data in freebase. This means that if, for example, the gene ontology were loaded into freebase on a particular day by a particular trusted authority, users of freebase could access versions of the GO topics exactly as they appeared at that moment. This answers the common worry from biologists about trust and authority. The 'oh no, I wouldn't use anything that could be edited by just anyone' phenomenon goes away and the best of both worlds becomes a possibility. With an open app like freebase you get the wikipedian potential for rapid, parallel, collaborative knowledge formation. With this kind of versioning, it should be possible to create trusted, authoritative versions at any point in time.

huge.

semantic dissonance in uniprot

2009-02-21T08:56:00.000-08:00

I was playing with RDF gathered from UniProt this morning and stumbled on something that made me a little sad. Apparently molecules like the rigor mortis (rig) protein from Drosophila (depicted at right) have been isolated from "a young sporophyte contained within a seed". This seems a little strange for a fruit fly protein.

In the RDF, I see triple 1: (Protein Q86BY9, isolated from , uniprot tissue 229 (Tissue Embryo))
http://purl.uniprot.org/uniprot/Q86BY9 (subject)
http://purl.uniprot.org/core/isolatedFrom (predicate)
http://www.uniprot.org/tissues/229 (object)

and triple 2: uniprot tissue 229, is the same as, plant ontology term PO:0009009 (embryo)
http://www.uniprot.org/tissues/229 (subject)
http://www.w3.org/2002/07/owl#sameAs (predicate)
http://purl.uniprot.org/po/0009009 (object)

Even if the sacrificial fruit flies were found inside some poor young sporophyte, this probably still would not make sense. I hate bashing Uniprot because the fact that they have bothered to produce RDF versions of their records is a useful and unusual trait for a major bioinformatics institute and has enabled a lot of my work, BUT.. RDF/etc. are really much more useful when some attention is paid to the definitions associated with their constructs. OWL:sameAs means that 'two URI references actually refer to the same thing: the individuals have the same "identity"'. Having live statements that say that a fly protein was isolated from "a young sporophyte contained within a seed" is probably not such a great thing.

Perhaps we would be better off if the W3C would introduce the relationship "OWL:kindOfSimilar" . Its likely that RDF/OWL providers like UniProt use OWL:sameAs in these situations because its the built-in predicate that comes closest to expressing what they mean. Maybe if we gave them something else to use in the many cases where similarity exists but is not absolute things on the semantic Web would make more sense. (Perhaps we would also be rewarded with standardized ways to quantify the confidence or degree of such relationships as well.)

Note to Uniprot: say what you mean!

Note to semantic Web people: make it easier for them to say 'squishy' things like "sort of similar to"! Reality is analogue, deal!

Notes from Everything is Miscellaneous

2009-02-12T13:36:00.000-08:00

I just finished David Weinberger’s book “Everything is Miscellaneous” and heartily recommend it. Appropriately enough, Amazon suggested it to me when I was looking for Clay Shirky’s book “Here Comes Everybody” which pretty much everyone in my online world has recommended and for Cass Sunstein’s book “Infotopia” which Michael Nielsen suggested to me in a response to a blog comment. Each of the three books is broadly about changes enabled by the Web, but they each focus on different aspects. To summarize full-length books on a variety of interesting topics in a couple words: Shirky focuses on the effects on human social groups, Sunstein on markets, and Weinberger on information itself. Of the three, I liked Weinberger’s the best – probably because it resonates most strongly with my own recent studies and interests. The topics cited in the book were uncannily related to my own personal reading/dreaming/experimenting list –Aristotle, Shirky, tagging, ontology, semantic Web, intertwingularity, blogging, Nature Publishing, Vanevar Bush, Dewey, it even mentions LSIDs of all things!

Weinberger writes in an imminently quotable style. Practically every paragraph contains a pithy sentence worthy of repetition. To illustrate, here are just a few that caught my attention (page numbers from paperback):

“...the solution to the overabundance of information is more information” p. 13

“...Making complex, meaningful phenomena explicit can leave us rudderless, force us to oversimplify, and result in statements that are incomplete and misleading.” p. 156

“The meaning of a particular thing is enabled by the web of implicit meanings we call the world” p. 170

“A Semantic Web that loosely stitches together imperfect, smushy, local efforts is not only more likely, it is to be preferred” p. 195

“Paper drives thought into our heads. The Web releases thoughts before they’re ready so we can work on them together.” p. 203

“That was Aristotle’s startling discovery: a thing, standing on its own, is what it is because of its connection to other things like it and other things not like it” p. 219

Obviously, I liked the book and tended to agree with most of its arguments (funny how liking and agreeing often go together). My only complaints about it related to its treatment of social tagging and of the semantic Web; these complaints likely arise from the book’s successful attempt to appeal to a very general audience and the coincidence that I might know a tiny bit more about these areas than the intended audience – hence I would like more detailed treatment of the consequences of more specific aspects of the technologies than is provided. For example, it annoys me when people talk about social tagging as if it is a permanent, stable technology, thus assuming that the ambiguity of tags (as simple Strings unlinked to concept definitions) is a fundamental aspect of all such systems rather than an optional weakness of the design of particular instantiations. It also annoys me when people talk about RDF as an “ontology language” – especially without at least making some attempt to explain its relationship to things that are most definitely ontology languages like OWL..

Overall a good quick read and a useful reference for anyone interested in the continuing evolution of the Web and, through it, our species.

non-anonymous peer review

2009-02-09T16:14:00.000-08:00

I spent this afternoon acting as a voluntarily non-anonymous peer reviewer - its scary. I ended up advocating rejection of the article I was reading and I have to say that Vince Smith(see end of linked post) was absolutely right that the act of signing your review "keeps you in check". Knowing from the outset that your words are going to be linked to your name can really change what you have to say - it certainly makes you think about it for a while longer. It is scary though - I hope that I managed to convey enough of my reasoning and suggestions for ways to improve the article that the authors don't despise me and attempt to ruin my life... I also hope that the editors of the journal manage to acquire at least one additional reviewer for this manuscript - safety in numbers! Or perhaps the editors will strip my name from my comments? Time will tell I guess.

Branche le monde

2009-02-03T09:15:00.001-08:00

Branche le monde...

On the right is a depiction of Francois Belleau of bio2rdf "linking the world" by Jean Giraud (AKA 'Moebius'). I love the sound and the meaning of this statement (which I found in the signature lines of Francois's emails). I liked it so much in fact that I used it as both the opening and the closing quote for my PhD dissertation. To me, it sounds both warm and powerful. As Francois explained to me (as I am language impoverished):

'"Branche le monde" means in French different things. Branche means to plug into, like to plug in an electrical wall socket, so branche means to connect into the Internet, someone, like me, is said to be "branched" or connected if you prefer. A branche have also the "tree branches" meaning in french. "Le monde" means also two things: the people and the world. So branche le monde means : Connect the people to the Internet

Plug the world to Internet but it could also means plug th people to the internet plugged to the world. or else

connect people together !'

Like I said, both warm and powerful, I love it!

p.s. the dissertation goes to the external examiner today - hope they like the quote ;)

getting attention

2009-01-23T13:39:00.000-08:00

I sent the first complete draft of my dissertation to my committee last week. Yay!

At my last meeting with them they promised a quick response if I got it to them that day - even saying that they would reserve time specifically for reading it. Unsurprisingly, most have not responded as they had promised - in fact most have not said anything. Boo! This means that either I will submit the thesis without having the benefit of their comments or I will miss my window for graduating this spring - double boo.

As many of you are no doubt aware, this is not an atypical situation. Whether it is a 200 page thesis document, a 2 page report for the local hospital, or a technical article being prepared for publication, it can be very difficult to get essential feedback. As Michael Nielsen has succinctly put it, the critical resource in science today is the attention of scientists. My committee isn't ignoring me out of spite, they are just unbelievably overworked people with limited amounts of attention to meter out and I happen to be lower on their personal totem poles than their grant applications, their manuscripts, their other students, their husbands, their wives, (and probably their dogs and cats).

The question this brings up is, when its critical that you receive some attention from a scientist or two, how do you go about getting it? Lets make this more specific and ask, when you need some one to review one of your papers (prior-to, or rather-than a journal), how do you go about acquiring that needed attention?

The only way that I have addressed this problem is by asking friends and family for help. This works well (depending on your friends and family) up to a certain extent, but has some significant shortcomings:

they like you and don't want to make you feel bad - which may influence their assessment of your work
they may reside within the same information cocoon that you do - which means they may have little additional knowledge to contribute
they eventually get tired of helping you out because they have their own problems to deal with

What other options are there? I suggest two that both revolve around markets. The first market exists and the second is yet to be created.

$Market #1$

In discussing whether it would be worth my time to take a free scientific writing course offered at my institute, a professor who had taken the course really encouraged me to take it. When I asked what she got out of it, she said that the most important thing that she learned was the value of working with a professional editor. She now pays an editor to review every research paper and every grant that she submits. I found that a little strange. The most valuable product of a class is to learn that you need to pay some one to help you do what the class was trying to teach you? Weird. I would have dropped it there, but in another conversation with a very talented and well respected author, the same advice appeared. To write at a professional level, getting professional help appears to be a vital component.

Now, as a student or a post-doc making lets just say not a lot of money, this advice is about as valuable as another suggestion that I love to hear from friends that actually have savings and real jobs (or rich parents), "oh, you should really try to buy a house, its such an important investment and now is such a great time to buy". Great, thanks. As soon as my scholarship check comes in I'll head out to the real estate agent... Lacking funds to actually pay money to an editor, what could I possibly provide in exchange for some scientific attention?

Market #2

Well, according to some definitions, you might actually call me a scientist. In fact, several journals have successfully taken my scientific attention from me (without any form of compensation) and handed it out to other scientists in the form of peer reviews. Maybe I could claim greater control over this process? Maybe there is a way to generate a market within which I could pay for attention when I needed it with my attention at other times. I'm not referring to the perhaps more exciting 'collaboration markets' that Dr. Nielsen discusses, at least not yet, I'm simply referring to a market for the direct exchange of literary review in scientific contexts.

Here is the essence of the deal; I will exchange my attention in reading and commenting on your paper in exchange for your attention on mine.

Here are some of the additional complexities that might make an implementation of this idea interesting;

the chance to accumulate 'reviewer points' so that the system could go beyond barter and towards a more complete kind of market
the opportunity for anonymity for authors and reviewers to ensure that you can always say what you think you should say
the opportunity for the lack anonymity - for reviewers to be acknowledged in future iterations of the work that they review
the chance for participants in the system to establish levels of trust - some reviews really are more valuable than others and this should be recognized

I think something like this is vital. It opens up a wide range of new opportunities for improving the way science works. Papers could be 'published' within this system and gradually accumulate findability-enhancing credibility (and improvements). Such assessments of credibility could be used to form a continuous rating scale for 'publications' that would replace the unnecessarily binary nature of journal-based publishing without losing the filtering effect touted by its proponents.

Such a system might improve on preprint archives like Nature Precedings by both providing a direct incentive for scientists to contribute comments on papers (most papers are never commented on at all at the moment) and providing a very direct approach to the filtering problem.

If anyone is interested in creating something along these lines, let me know. I hope that I will be needing a job soon.

p.s. Thanks to Mikele Pasin for thoughts we shared on the 'Paper Demolisher' at KCAP 2007 that are directly related to this post.

the madman

2009-01-08T15:37:00.000-08:00

In the classic book, The professor and the madman: A Tale of Murder, Insanity, and the Making of The Oxford English Dictionary , Simon Winchester tells one of the first stories of truly successful mass collaboration. In the story, Winchester describes how one man, who it turned out was quite mad, contributed a disproportionately large quantity of the definitions used to create it.

I've seen this pattern repeated many times in my studies, though the figures on the leading edge of the charts generally don't tend to be crazy. Its fascinating how consistent the pattern describing the volume of contributions produced by individuals in open settings is, a few people always end up doing most of the work. I was wondering, does anyone have any data about the actual fraction of definitions produced by different individuals for the original OED? It would make an awesome figure for any text about volunteer based mass collaboration (such as the dissertation I should be writing right now).

I was wrong - 2008

2009-01-01T19:06:00.000-08:00

I’ve been thinking a little bit about how to finish 2008 and start 2009 here on i9606 and the things that spring first to mind are mistakes made in my thinking this past year. Maybe this is because mistakes are where learning happens or maybe its simply because I tend to be self-critical in my self reflection. In any case, here are just two of the many things I got wrong this past year with a few thoughts on why.

I really thought I was going to graduate in 2008, but I was wrong.

I think that this happened for two reasons, first, I got distracted by a side-project and second, I didn’t have a clear idea of what really constitutes a dissertation. In hindsight, when the project that was supposed to take two weeks reached, maybe 5 weeks, I should have dropped it and moved on – well before it had a chance to eat up seven more months of effort. I think that if I had a clearer vision of the end goal, of the dissertation in this case, I might have made better decisions regarding the use of my time in the short term.

I wrote in a comment in a post by Timo Hannay that I would continue to blog even if no one ever read my posts, that the most important audience for these words is my future self rather than friends and unknown others, but, once again, I was wrong.

At the time, I was reacting against the apparently popular desire to bottle up the blog form, especially in scientific contexts, into something overly specific, to classify it as science writing for a popular or otherwise uninformed audience. To me it never really felt like that. It has always been an intensely personal platform without any higher purpose beyond myself and I always considered my audience likley to be better informed than me. I confess. I write here exclusively for my own benefit and believe that most other writers do the same. Though there are a few others, the main reason I started to write here and still do is to attract feedback on my ideas. I am after that most precious resource, the attention of informed and intelligent colleagues. Without occasional comments, I suppose that I would eventually give it up and move things over to a private journal. So, thanks everyone for all of your responses in 2009, I really appreciate them and will do my best to reciprocate where I can. (The consequences of this need for feedback and the pattern of seeking it in public, digital space are, I believe, interesting and wide-reaching, but that is food for another post.)

What mistakes did you make in 2008?

Happy 2009 everyone!

oranges on the beach

2008-12-18T21:24:00.001-08:00

I feel I should give due credit to my friend Perseus who, several years ago, suggested the "selling oranges on the beach in Jamaica" alternative lifestyle option I made use of in my last post.

As you may have noticed from a few of my posts here, I, like many other grad students, have the tendency to grow depressed and feel sorry for myself from time to time. In two cases during my time here as a student, acquaintances have literally died as a result of their failure to deal with this phenomenon (and likely other problems in their lives). In reaction to these tragic events, Perseus suggested that you should always consider the worst possible scenario - you fail your comprehensive, your girlfriend leaves you, you are deported, you are fired from your part time job at Wendy's - and realize that its actually not really that bad; there are still options available that would result in happiness, like finding your way to the tropics and figuring out how to make tourists + local fruit = enough money to live on till the next day.

Whenever the pressure seems too great, my work seems too trivial, or I feel like some one is finally going to catch on that I don't have the talents needed to continue on in my current endeavors, I try to think of that. If all goes to hell, I will simply point my way towards warm weather and sort something out - it might even turn out to be more fun anyway!

an iPhone future?

2008-12-13T16:48:00.000-08:00

Like many of my friends, I am finally coming close to taking the next step in my professional life (or so I like to tell myself) and, of course, I am wondering what direction to take. If I had more of a desire to move to London, I'd love to try to work for Nature Web Publishing, but that seems unlikely at the moment. One of the other possibilities is to go into the family business and help my father build apps for the iPhone like Relations, his first foray into the field.

Here is the obligatory family plug, Relations is a cool new way to organize, navigate, and show off your family pictures on your iPhone. Try it, you will like it! Its even semi-semantic :) - the links between the pictures now have added meaning...

What do you think? Join the family business? Try to continue on in science somehow? Join another startup (like maybe this one)? Start up something on my own? Apply for work at Freebase? Sell oranges on the beach in Jamaica?

UMLS semantic type clouds

2008-12-06T17:23:00.000-08:00

I'm trying to understand the relationship between tags used to describe academic citations by researchers and the descriptors used to describe the same citations by the pros. To do so, I've been trying to get a grasp of the differences at a level up from strings using tools from Unified Medical Language System. Right now I'm having my doubts about whether this is really going anywhere. There are differences, to be sure, but I'm not convinced right now that they are really meaningful. If the data you see below says something to you, I'd be obliged if you could tell me what it said. You should be able to access it in raw form in my newly generated area of the Many Eyes website.

The following tag clouds were generated for the terms used to describe about 87,000 Pubmed citations using Many Eyes. Each tag in the cloud corresponds to a UMLS semantic type. For the first cloud, the types are related to the MeSH descriptors associated with the citations. For the second cloud, the types are related to the tags added by Connotea users to the same citations. There is some bias here for the Connotea tags because I only considered tags for which I could rapidly, computationally identify a matching UMLS concept with high precision - so many of the tags aren't actually represented in the cloud. 'Functional concept' seems to be a standout for Connotea. It represents "a concept which is of interest because it pertains to the carrying out of a process or activity". There is a wordle with the Connotea tags that linked to functional concepts at the bottom.

Connotea tags below, MeSH terms above

Connotea tags for functional concepts below.

boundaries

2008-12-04T20:21:00.000-08:00

Of networks, classes, hierarchy, groups, and the major transitions in evolution.
(warning, not-very-well-thought-out stream of consciousness to follow)

Like a lot of people, I'm currently reading Clay Shirky's book 'Here Comes Everybody'. One of the dominant themes of the book is the formation of groups of people, in particular groups of people that appear to rapidly self-assemble on the Web and produce some interesting product or behavior. This idea of the auto-genesis of complex systems is taking me back to about 9 years ago (yikes) to a time when I was thinking a lot about evolution. Way back then, I was trying to think up better ways to intentionally evolve teams of independent units (e.g. robot control systems) that had to interoperate with one another to achieve some group purpose (e.g. win robocup). My thinking (and a great paper about chickens) lead me in the direction of multi-level selection theory, but that is another story. Among the other people around me at the time that were also thinking a lot about evolution, was one Pietro Speroni.

Pietro was working on a new Artificial Chemistry (a subject I know next to nothing about) with which he was exploring biologically-inspired mechanisms for automatically generating increasingly complex systems. The key aspect of his work that has been trying to creep up to the surface of my consciousness since I started with this Shirky book was a focus on defining his simulated molecules such that they could link together to auto-generate boundaries akin to cell membranes. This was considered the key simulation behavior of interest because the genesis of each new boundary, e.g. the boundary that defines an organelle in a cell, might be considered another step up the ladder of complexity. With each new boundary, a new kind of thing emerges in the system with which the other things can begin to interact with in new ways and thus complexity can increase. Systems like this are interesting because one such system appears to have generated the most fascinating thing imaginable, us. The fuzzy connection dangling around here is between the evolution of complexity in the abstract as Pietro was studying, and the very real evolution of complex groups within human society that is being made possible by the social Web.

Each new user-generated node in the network, each blog posted, each tag applied, each photo shared - each captured communication - can be seen to either implicitly or explicitly define a new group of people. Such groups could contain the people that used the same tag in Del.icio.us, who tagged the same website, who 'liked' the same photo etc. Groups could also contain 'friends' of people who tagged a particular photo or people who tagged a website that was tagged by other people who tagged another website ad infitum. Within the increasingly large and multifaceted network we are (most often accidentally) creating via our contributions to the Web, human groups of seemingly limitless complexity already exist and this trend seems only to be increasing. Do these groups have any relationship to the groups of molecules defined via physical relationships that Pietro's model was emulating? Is there anything fundamental going on here?

I don't know. Just had to get that out of my head.

scifoo mailing list madness

2008-12-02T09:59:00.001-08:00

The folks at O'Reilly have started up a mailing list for scifoo alumni and seeded the conversation with the question "Who's doing the coolest, most interesting work in emerging technology?"

The desire to reproduce the conversation here for you is powerful but I feel constrained against simply copying and pasting the messages as its still a private mailing list (though I doubt that any of the people posting would care).

Hopefully when I'm finished writing my dissertation I'll be able to dive into some of the threads in there a little bit and share them with you (at least as wordles).

coverage of Pubmed by Citeulike

2008-11-17T10:29:00.000-08:00

In the life science domain, the total number of items described by social tagging systems is currently tiny in comparison to the number of resources described by institutions. To illustrate, the MEDLINE bibliographic database contains over 16 million references while, as of November 9, 2008, Citeulike, the largest of the academic social tagging services, contained references to only about 203,314 documents with known PubMed identifiers.

Though important to see where things stand today, the interesting aspect of these new systems right now is their potential for growth. Given the large numbers of contributors (and very large numbers of potential contributors), it seems possible that their coverage might eventually meet or surpass that of resource-contrained institutional mechanisms. In 2007, the NLM reported that it indexed 670,943 citations for the MEDLINE database which equates, on average , to about 56,000 citations per month. To estimate if social tagging services might someday reach the same level of throughput as the NLM indexing service, we compared the rates of growth, per month, for MEDLINE and for Citeulike (on Pubmed citations) over the last several years and used this data to make some predictions of future trends. Here is what we came up with.

The figure plots the numbers of distinct Pubmed citations described by users of Citeulike and by NLM indexers each month and, using exponential smoothing, plots an extrapolation of the observed trends several years into the future. Based on the data obtained so far, we find that the numbers of biomedical resources described per month by Citeulike users is increasing more rapidly than the the number indexed per month by MEDLINE and that, if current trends continue, Citeulike coverage would catch up with MEDLINE around the year 2014 - at which point both systems would be describing approximately 70,000 biomedical citations per month. As the rapidly expanding confidence intervals illustrate, there is insufficent data to provide strong evidence for the precise point of intersection or even that Citeulike will continue to grow; however, it seems plausible that Citeulike and other scientifically oriented social tagging services will continue to expand in their coverage of the life sciences domain at a faster rate than institutional systems and thus will eventually catch up to the point where every document indexed by a professional is also tagged for personal use by a scientist (or 10).

So, what are we going to do with all of that data ?

current interest levels in scientific social tagging

2008-11-14T14:57:00.000-08:00

Lately, I've been looking at the data emerging from social tagging services in the academic domain, particularly Citeulike, Connotea, and Bibsonomy. Here is how these three services look in terms of number of unique visitors per month for the past year.

It seems fairly impressive until you compare it to say, NCBI...

At which point, they seem either a) pitifully insignificant or b) primed with the opportunity for dramatic growth.. If even a small fraction of the people that visit NCBI started to use these kinds of services, the data they could generate (e.g. simply metadata in the form of tags added to scientifically relevant things) is massive.

presidential election 2008

2008-11-04T23:10:00.000-08:00

Canada's response to the results...

WooHoo! Lets get wasted!!

dullhunk PloS Bio Article

2008-11-01T15:31:00.000-07:00

Duncan Hull and company have just published a thorough (210 references..) review of the current state of scientific digital libraries. Anyone interested in the changing face of publishing or of the Web in general would likely find the article interesting. Out of the many ideas discussed, these two caught my attention:

"As we move in biology from a focus on hypothesis-driven to data-driven science, it is increasingly recognized that databases, software models, and instrumentation are the scientific output, rather than the conventional and more discursive descriptions of experiments and their results."

and

"We suggest that the main obstacles to warmer libraries are primarily social rather than technical in nature. Identity, trust, and privacy are all potential stumbling blocks to better libraries in the future."

To the first quote, I will say simply, hear hear! The idea that the units with which scientific progress is published and thus measured should correspond more directly to discrete, integratable chunks of knowledge and to sharable processes for knowledge generation rather than (often unparsable) stories is one whose time has clearly come.

The second one stuck out because it is so similar to something I played a part in writing a while back. In another review article (with a paltry 93 references) we suggested that ".. the primary hindrances to the creation of the SWLS [Semantic Web for Life Sciences] may be social rather than technological in nature ..". In a sense, the SWLS that we were thinking about way back then could be said to subsume the digital libraries of Dr. Hull et al.'s article. But... looking more closely at the first quote above, I realize that relation isn't subsumption, but equivalency. Though they are writing specifically about 'libraries', they clearly consider databases, software, etc. as parts of the new incarnations of libraries and thus are writing about exactly the same thing that we were writing about.

Its interesting that we came to similar conclusions to some extent, both articles suggest that the main challenges in moving science forward on the Web are in handling problems that have people at their center, not technology.

BioMoby Annotation Jamboree

2008-10-22T20:31:00.000-07:00

If you happen to be a BioMoby developer, a bioinformatician, or just interested in helping us out and possibly making $20 of gift certificate remuneration, come have a go at annotating some BioMoby web services with a specialized version of the Entity Describer. You will need to use Firefox, Safari, or Opera - no Explorer support so far.. and jam on over to the jamboree website.

Warning, it is seriously difficult..

We've been working hard to produce a system that makes this task (a) possible and (b) as intuitive as possible. However, even when you know what you are doing and have figured out how to use the interface, the task of describing things you don't necessarily know intimately with a vocabulary you are also unfamiliar with is a real challenge. I'm very curious to see what people end up doing and to assess how well we can pull their collective efforts together.

Working on this project has reminded me just how difficult interface design (and implementation in javascript) really is. It is absolutely stunningly surprising to watch people try to use a tool like this for the first time. They simply never do what you expect them to, nearly everyone seems to react differently, and no one reads the instructions (even while I stand right beside them encouraging them to do so). The challenge of building software to aid in accomplishing a task that most users will never have done before to satisfy a purpose they ubiquitously have difficulty grasping has been quite an adventure in iteration. We think we have something that works, finally, but we won't know until you give it a try now will we ;).

Heading to ASIST

2008-10-22T19:55:00.000-07:00

This Saturday I will be heading down to Columbus, Ohio to attend the annual meeting of the American Society for Information Technology. If you are coming, hope to see you there!

I will be presenting a poster about an empirical approach to the study of (human) indexing systems based on the comparative analysis of the syntax of the terms that are used within the different systems. The work emerged from investigations of the relationship between social tagging (e.g. Connotea) and professional indexing (e.g. MEDLINE). As has been typical in my career thus far, when I got started on the project, I found that I didn't have the tools I felt that I needed to answer the question satisfactorily so I spent most of my time trying to make them.

One of the more interesting outcomes of the work are visualizations of the differences between the structures of the terms used within ontologies, thesauri, and folksonomies. They are surprisingly distinct from one another. The abstract for a full paper (just accepted today) describing the work as well as the programs written and the data processed are available here.

update to social semantic tagging list

2008-09-29T11:33:00.000-07:00

A while back I started to keep track of social semantic tagging projects (other than mine) in a post. Since then, I've continued to add to the list in that post - most recently today. Its starting to feel downright crowded in here - not the feeling you want to have when trying to wrap up the last project for your dissertation.

fret fret

headphones

2008-09-24T20:46:00.000-07:00

This morning I decided to stay home to concentrate on the new and surprisingly involved task of designing and coding up the pattern for subject-task assignment for an upcoming experiment such that the paltry number of subjects I'm likely to be able to recruit will provide me with that magical value of the P family sought after by so many other desperate graduate students like wolves after a wounded moose. (Wow, check out the length of that intro sentence!). Anyway, as I settled down to think through my design, I was plagued by one auditory distraction after another - car alarms, truck engines, my neighbor's bass, even jackhammers (I kid you not). Growingly increasingly frustrated with the noise and my inability to concentrate, I realized that the solution was not less noise, but more noise! I strapped on my headphones and went back to work. In 5 minutes I was not only working more effectively, I was calmer and frankly happier.

I've heard many people, usually a few years my superior, complain about people listening to music while they 'should be working'. I wonder if this reaction has anything to do with the fact that they grew up in a time when it was actually possible to find a quiet place to work on a regular basis? Now, in increasingly compressed living and working quarters, it seems that music is one of the few ways left to provide oneself with a bit of cognitive privacy.

Do you listen while you work? What is your favorite artist/genre for getting something done?

one more sci foo tag cloud

2008-09-13T00:03:00.000-07:00

Alright, I know this is getting a bit old now, but I read this summative post about the meeting by Frank Wilczek (I got the first signed copy of his new book ;) and thought it would make a much more descriptive cloud than the ones I put together during the meeting about the attendees and their ideas for sessions. So, here is yet another wordle-generated scifoo cloud based on Frank's post. (It is better if you click through to the larger image).

freebase ED and sparql

2008-09-12T12:31:00.000-07:00

So what do you do when the two papers you would like to finish and submit are sitting in the hands of co-authors? Kayaking? Sleeping? Surfing? today, no. Hacking? today, yes.

While I wait, I decided to finally start working on bridges between freebase and our semantic tagging repository for ED for use after the data is collected. To get started, I wrote the code to answer this question: "what URIs have been tagged with the organism classification X or any of the sub classifications of X". For example, has anyone tagged anything with magnoliopsida or any of its lower classifications, such as arabidopsis?

To do this, I needed to utilize the 'Higher classifications' (or of course 'Lower classifications') property of the Organism Classification Type. Unfortunately, there is, thus far, no such thing as a generic transitive property in freebase as far as I can tell, so I built a brute force, recursive query that implements it myself. I send the following with the '???' replaced with my starting point (e.g. 'magnoliopsida') to freebase as the rest of the URL starting with http://www.freebase.com/api/service/mqlread?queries= .

{"q1":{"query":
 [{
 "higher_classification" : "???",
 "name" : null,
 "guid" : null,
 "type" : "/biology/organism_classification"
 }]
}}

Freebase responds with the lower classifications of my query and then I repeat the process with these until either a maximum depth is reached or it bottoms out. If you know a better way to do this please let me know.

Once I have all of the guids for all of the lower classifications of my query, I send these over to get URIs tagged with any of them via a SPARQL query like this:

prefix tag: 
prefix rdfs: 
select ?tagging ?tag
where {
{
?tagging tag:associatedTag ?tag .
?tag rdfs:isDefinedBy http://www.freebase.com/view/guid/9202a8c04000641f8000000000516f8d
}
UNION
{
?tagging tag:associatedTag ?tag .
?tag rdfs:isDefinedBy http://www.freebase.com/view/guid/9202a8c04000641f800000000003be00
}
UNION
{?
tagging tag:associatedTag ?tag .
?tag rdfs:isDefinedBy http://www.freebase.com/view/guid/9202a8c04000641f800000000572e4660
}
}

The query has as many UNIONs as topics to check for. (Note that you have to put URIs in SPARQL queries inside angle brackets - blogger was making this difficult for me to include). It works well enough, but if there are too many, I hit the max URL size limit (HTTP 414) so I set it up to send them in chunks and then reassemble the results.

Hacky? Yes. Successful for demo purposes? so far..

Any ideas about optimizing such activities most appreciated.

On the todo list:

Assemble the must-tag list of web services for the upcoming biomoby/ED jamboree
Build up an API-like library of queries like the above and normal queries like 'get all the URIs tagged by user X' so that we can more easily put up reasonable human interfaces for users of ED2.0. (Thanks to those that have already started using it!). Note that any developers out there already have access to all of the data needed to build ED applications via HTTP calls to freebase and to our repository. The library I speak of will be used by us and probably made public, but the real idea is for external developers to utilize SPARQL/MQL directly as that provides the most flexibility.
Create mappings between bio-ontology classes and freebase topics. Likely follow Shawn Simister's model for approaching this integration. (He has some excellent ideas about SPARQL/MQL integration).
Prepare for kayaking trip tomorrow
Graduate before they cut off my funding...

fear and loathing in academentia

2008-09-07T15:26:00.000-07:00

OK, I'm mad and probably shouldn't write the following today. Oh well.

Now, here is why I am mad. I've had a paper rejected by the Journal of Biomedical Informatics on the basis of one review. It took two months to get this review. The review does not seem fair and certainly does not provide useful guidance about how to improve the quality of the science described in the paper. Here is JBI's response in its totality, with some embedded reactions from me in red.

Ms. No.: JBI-08-163
Title: OntoLoki: an automatic, instance-based method for the evaluation of biological ontologies on the semantic Web
Corresponding Author: Dr. Mark Denis Wilkinson
Authors: Benjamin M Good; Gavin Ha; Chi Kin Ho;

Dear Dr. Wilkinson,

I requested that my advisor be the corresponding author on the paper because I would be traveling right after the submission and am hoping to relocate soon.

Experts in the field have now reviewed your paper, referenced above. Based on their comments, we regret to inform you that we are unable to accept your manuscript for publication in the Journal of Biomedical Informatics.

We have attached the reviewers' comments below to help you to understand the basis for our decision. We hope that their thoughtful comments will help you in future submissions to the JBI and in your future studies.

Sincerely,

Janine Burch
Journal of Biomedical Informatics, Editorial Office

Elsevier
525 B Street, Suite 1900
San Diego, CA 92101-4495
USA
Phone: (619) 699-6392
Fax: (619) 699-6211
E-mail: jbi@elsevier.com

Reviewers' comments:

Reviewer #2:

Its a little odd that we only got to see Reviewer 2's comments. I don't know if anyone else reviewed it or not.

Good and his colleagues present OntoLoki, a very interesting approach for data-driven ontology evaluation. The novel idea is that the quality of ontologies can be measured automatically although ontologies without or with very few formal restrictions on class membership are used. For poly-hierarchically organized classes suitable datasets with positive examples - i.e. instances with properties - as well as negative examples are composed. Machine learning algorithms are used to determine empirically those rules (patterns of properties) that allow predicting class membership reliably. With other words: The ideal situation is to find instances of classes like "Cat" with properties like "furry" allowing their consistent assignment to the class "Cat" and discrimination to neighbour classes like "Bird".

Yep, that pretty much sums up the general idea. So far, so fair.

There are a lot of inherent challenges with this approach that are addressed by the authors, e.g.
- the dependence on the context (chapter 1) and on the way of determining instances and their properties (chapter 1.2.1)
- the problem of sufficient number of instances for every class for estimating a class predictor (chapter 1.2.4)

These are common problems when using empirical approaches. However, the reviewer has doubts about the suitability of the OntoLoki approach for evaluating ontologies. The authors themselves admit that especially the results of the Cellular Component experiment are suboptimal, see chapter 3.2.1 (only 17% is evaluated) and chapter 3.2.3 (results are not overwhelmingly illuminating).

OK. At this point the reviewer has pointed out that we correctly identified challenges with empirical approaches to ontology evaluation and discussed them in respect to our approach in the paper. Both of these challenges, context-sensitivity and data dependency, are fundamental to any methodology that is based on the use of data to help answer a question. Keep in mind that the main point of the paper is to describe and evaluate a method. To do so, we explain it and then test it out in a variety of different scenarios (different ontologies and different datasets). In some cases it is successful and others it is not. By describing the results from all of these experiments we faithfully represent the realities of applying the method.

The reviewer criticizes the method by pointing out our own admissions regarding problems encountered with the dataset assembled for the evaluation of the cellular component branch of the gene ontology without actually saying anything about the method itself. Perhaps, criticism could fairly be placed on our data collection methods for that particular ontology. However, the point was not to evaluate that ontology it was to evaluate the proposed method. That 17% number resulted because we didn't collect enough instances to evaluate the other classes. If we collected more data, the number would have been higher, but that is completely irrelevant to the utility of the method and our evaluation of it. In fact, by including data like that, we much more accurately present both the positive and negative aspects of the method. Perhaps next time we should simply obscure any negatives to avoid such criticism.

The MAIN PROBLEM the reviewer has with this approach:
OntoLoki tries to solve a structural classification problem empirically that originates in poor defined ontologies. Instead of (suboptimally) trying to determine the consistency of ontologies with no formally defined restrictions on class membership the ontologies should be enriched by such formal definitions on class membership, see http://bioinformatics.oxfordjournals.org/cgi/content/abstract/22/14/e530.

Now, here is where this starts to get ridiculous. "the ontologies should be enriched by such formal definitions". Well, we couldn't agree more! That is one of the main reasons we did this! OntoLoki provides a starting point for doing just that!

As the reviewer seemed to understand in the summary of the paper above, the method is intended to be applied to ontologies (or whatever you want to call class polyhierarchies used in classification situations) that aren't necessarily formally defined. The rules that are learned could be used to suggest possibilities for formal class restrictions that are based on the data the classes are already associated with.

As a matter of fact, the OntoLoki method can actually be used on formally defined ontologies to identify candidate expansions of other definitions. We recognize the importance of these definitions and the reasoning they allow for, that is why the reference so generously provided above was one of the main citations in the paper! In fact, the ontology described in that paper was used as a benchmark of quality for other ontologies - thus providing us with a means to evaluate our method.

In the introduction Jeremy Rogers and later on Barry Smith are referenced as proponents of ontology evaluation. However, these and other researchers in the field of biomedical ontology are mostly concerned with the quality of explicit formal definitions and the structure of ontologies, see http://ontology.buffalo.edu/evaulation.html.

What exactly does the "however" mean here? Indeed, both of these scholars are involved in ontology evaluation and I would say are, in fact, proponents of the idea. Why the contrasting "however"? The quality of formal definitions and the structure of ontologies (which can of course result directly through inference applied to those formal definitions) are certainly aspects of relevance to the domain of ontology evaluation.

The structure of ontologies must certainly have something to do with the inferred or asserted class hierarchies they produce. The OntoLoki method is designed for evaluating these hierarchies. So.. why this statement ? Perhaps you could argue that the method is not useful in achieving the task, but it doesn't make any sense to say that the task is irrelevant as seems to be implied here.

The very idea of ontologies is to have explicit criteria for deciding class membership of instances opposed to ambiguous language terms denoting those classes. If there are artefacts with no formally defined restrictions they should not be called ontology.

Alright, now we've got to essence of this so-called "review". The reviewer doesn't believe that the things the method was built to evaluate should be called ontologies. So they don't believe the Gene Ontology is an ontology and they don't believe that most of the ontologies in the OBO foundry are ontologies. OK, fine. Perhaps the reviewer should have suggested that we change the title and used a different word to describe whatever it is these things are. The complaint has absolutely nothing to do with the manuscript! The maddening thing is that we have been (sometimes very lonely) proponents of the expanded use of axiomitized, property-based definitions in biological ontologies for years and are still very much of this view. To be criticized for the community's fairly slow uptake of these methods makes my head feel like its going to explode.

However, the machine learning methods are very interesting for supporting different purposes in the context of REAL ontologies WITH formal restrictions on class membership", see chapter "Making use of OntoLoki" in the discussion section. The whole paper should be rewritten oriented to those other supporting purposes in the context of developing, using and evaluating ontologies.

Well thanks. It seems that some of the applications of the method (and the software we developed) are "very interesting" but only in the context of "REAL ONTOLOGIES". As it turns out, the method and implemented code could be applied directly to REAL ONTOLOGIES without alteration. (Note that the capitalization is from the reviewer).

This paper, submitted as a paper in the Biomedical Informatics Journal, is a copy of a Technical report, see http://bioinfo.icapture.ubc.ca/bgood/OntoLoki_14.pdf.

That this is even mentioned as a presumed negative is outrageous. The report (which does in fact contain the same content as the submission) is not a peer-reviewed publication, it is simply a very informal pre-print. Posting it is perfectly in accordance with Elsevier's rules when it comes to pre-prints, rules that it is clear the reviewer is not aware of.

It is far to long and should conform to the editorial guidelines of the journal.

First, I actually agree that it is probably a bit too long. We discussed this at some length before deciding to submit the full version and, in the end, decided that the length was warranted in this case in order to present the argument and experiments in completion. We could shorten it, and likely will when we resubmit to a different (open-access) journal, but that was actually one of the reasons we chose JBI - they explicitly state that there is no "arbitrary limit on the length of individual articles". The submission was well within the editorial guidelines of the journal - guidelines which the reviewer, again, was clearly not familiar with.

Ok, my rant is over now, the red has drained out of my face and I can no longer hear my heart beating in my ears, so I will switch back out of the red to conclude.

So, Reviewer #2, who are you?

One of the more impressive people I met at SciFoo told me that he has been signing his reviews for years to "keep himself in check". If reviewers had to sign their reviews it seems that perhaps they might be forced to do a better job. Good quality reviews (either arguing for reject or accept) would provide another form of publication - another way for scientists to get credit for the work that they do. Are you up to it? Sign your next review.

Peter (Google) and Christine (the librarians)

2008-08-31T17:24:00.000-07:00

I was very lucky to have my new wife with me at SciFoo for many reasons, not the least of which is that she is much better at socializing than I am and thus managed to introduce me to many people I would never normally have met. One of those people was Christine Borgman, Professor & Presidential Chair in Information Studies at UCLA. While I was struggling to explain my work on the Entity Describer project to her, she noticed Peter Norvig walk by and dragged him over to join our conversation. I guess she must have known him from somewhere but I'm not sure where. Anyway, I didn't realize this at the time, but Peter is head of research at Google. Ahem... did I mention that SciFoo interactions could be somewhat intimidating? So there I am, standing between two giants of modern information science trying to explain what it was I was doing there but mostly trying to get some insight into their respective thoughts on the organization of the world's information. It wasn't a long discussion, but here are the basics.

The main question that I posed to them was whether or not and how semantic tagging (a la ED) is or might be useful. On the surface, the answer from Peter was no and the answer from Christine was yes. However, the truth of the matter is that they were really talking about supporting different functions for the end user - though this fundamental difference became a little lost during the conversation. Google is principally focused on providing the best possible results, to the most people, given the least amount of information in the query - that is, keyword based search of the entire Web. Library-science is typically much more concerned with providing the capacity for people to make very specific requests using much more sophisticated queries that operate over much smaller collections of information (e.g. the library of congress). The fact that there is some overlap in the information needs of the users of these different kinds of systems often brings up the desire for combative comparison, but I think that, in reality, there is clearly no need for combat because they are simply too different in the functions that they intend to provide.

My interpretation is that Google isn't really concerned with intentionally provided meta-data in the name of end-user, full-Web search because the scale that they operate on seems to render any such indexing by one or even a number of parties almost laughably shallow in its characterization of both the nature of any particular item and its expected relevance to a query. When you have literally millions of people passively voting and indexing every item of the Web through their decisions to link to it or not, you have very sophisticated algorithms for understanding the text in the pages generating and receiving those links and to top it off you record and process millions of people's behavior when faced with your search results, why should you care what some person or institution says the item is about? The fact that they (among other search engines) beat out the directory-based approach to finding information on the Web is a clear demonstration that automatic indexing and link based relevance ranking do a better job than meta-data based classification - for the problem of Web scale search. Google clearly doesn't need human semantic indexers to succeed, though, as Peter said, they certainly use all of the information that exists. If there happen to be good indexes online (as Connotea turned out to be be for a fairly brief window), then their algorithms will certainly find them and use them - if not, no worries, the algorithms will take advantage of the 'normal' data on the Web and do just fine thank you.

From the library-science professional perspective, this attitude is clearly annoying. If human indexing isn't really necessary to find things, what is the point of the field that has devoted itself to the creation of effective ways for people to categorize things for retrieval? For example, there is a lot of annoyance that the Google Books initiative seems to ignore most, if not all of the meta-data already associated with the books that they are scanning and indexing. This means that meta-data, even as basic as volume numbers, is inaccessible for searching. For the library professional that is trained to both search through and construct careful and precise classification structures,the inability to even search for a specific volume of a book is infuriating - particularly knowing that it is well within Google's power to incorporate such abilities into their system.

So on the one side we have the perspective that there is still value in the careful, intentional use of meta-data in the search and retrieval process while on the other side we are quite happy to let the intersection of algorithm and massive passive indexing do the work. I guess, as is the usual answer, I'd suggest that both sides provide useful functions that are both worth keeping and advancing. A detailed classification system, either constructed intentionally through the work of professional labor or semi-intentionally through the work of social taggers, provides functionality that is clearly different than what can be achieved by automatic indexing; however, it may not provide any help whatsoever in improving a full-Web-scale keyword-based search. The essence of the power of intentional classification is the precision of the queries that it enables. For example, if I want only version 3 of "The Devil's Rights and the Redemption" and thats it or I want only those items that have been tagged as with bioinformatics and to_read by Jaa, there is really no way ( AFAIK) to accomplish this without the intentional recording and utilization of meta-data about those resources.

So, though Peter and Google may have little direct use for ED and its semantic meta-data generating and consuming brethren emanating from the library and information sciences, there are still clearly meaningful applications of such work. It just happens that providing effective search over the contents of the entire Web based on a string like 'Britney Spears' isn't really one of them.

I'm ok with that.

things I should write

2008-08-24T16:55:00.001-07:00

This always happens. I go to a meeting of some kind, get loaded up with new ideas and experiences, get excited about writing about them, and then reality comes crashing back down and I never end up finding the time to do the writing properly, if at all. Before the memories of scifoo slip further away, I just want to note some ideas I had for posts about the experience. If any of them strike your interest, let me know and I will do my best to do a complete post on the topic.

People I met at scifoo
Things I collected at scifoo
Google's role in the future of bioinformatics
Micro-credits. How to assign credit and blame in the smaller, newer bits that compose pieces of open science projects or other useful collaborative endeavors
My brief talk with Geoffrey Carr, the science editor for The Economist
Google could care less about meta-data, my talk with Peter Norvig and Christine Borgman.
What I would do differently if I ever go to another foo camp
Why I should have been twittering
cell phones versus one laptop per child
Impressions of the Google campus

I'll leave it at that, let me know if anything strikes your fancy!