nodalpoint.org - gene ontology - Comments

RDFs

cariaso — Fri, 11 Apr 2008 22:45:40 -0400

I came across this old thread, and thought I'd point out that these days SNPedia does have the sacred RDFs thanks to the semantic-mediawiki.org extension. Today the most obvious benefit is that the SNPedia:FAQ is able to maintain some live counts about the site. The RDFs aren't yet connected to any others, but one baby step at a time it looks a bit more semantic. It seems only a matter of time until it can leverage SPARQL and perhaps the Halo browser.

Cladistics

nuin — Tue, 08 Jan 2008 10:09:16 -0500

That's already a cult. A couple of more years it will be a completely new religion.

Religious War or Religious Cult?

Duncan — Tue, 08 Jan 2008 05:56:30 -0500

Theres lots of good stuff in the Open Lab 2007 winning entries but I particularly liked this: Academia is a [religious] cult. Sometimes the semantic web seems more like a religious cult than a scientific endeavour :)

Call for (more) Semantic Web Testimonials

Duncan — Mon, 10 Dec 2007 08:15:20 -0500

In addition to rather old and tired looking semantic web testimonials, Lee Feigenbaum is asking for more, see forwarded message below.

So if you've a SPARQL in your eye, these people would like to hear from you.

======

Dear W3C Colleagues,

I'm writing to you as a representative of an organization that is a W3C member whom I believe to be involved in research and/or development surrounding Semantic Web technologies. (My apologies if I'm mistaken.) In particular, I'm writing in reference to the SPARQL query language and protocol, which has recently been published as a W3C Proposed Recommendation, and is scheduled to be published as a W3C Recommendation in January.

In anticipation of sharing this achievement with the public, we are looking to gather a substantive collection of testimonials in support of SPARQL. I'm including below a note from Janet Daly of the W3C Communications Team which includes more details and guidelines about testimonials.

This is a great opportunity to help with outreach for a key semantic web technology to the rest of the world. At the same time, testimonials are an excellent vehicle for advertising your organization's support of W3C Semantic Web technologies as well as any products that your company might have that implement or use SPARQL.

The deadline for testimonials is Jan 8, 2008.

For more detailed information on the testimonial process, including examples, guidelines, and much more, please see Janet Daly's note below.

thanks,
Lee Feigenbaum
On behalf of the RDF Data Access Working Group

PS If you think you might provide a testimonial, I'd appreciate a heads-up note, just to help me track our status. Thanks!

~~ more information from Janet:

Dear DAWG participants,

Please accept my late congratulations for the set of SPARQL Proposed Recommendations.

In anticipation of a successful completion of the review period, W3C is preparing for announcements around SPARQL Recommendations, or standard. As part of the celebration, we will be issuing a press release and would like to include a testimonial sheet from W3C Member implementors and others.

We will have a draft press release available in the next week for your review. The document will be available to those with Member access. It will not be made public until the actual announcement date. Our current plans are to issue an announcement in early January; we're making efforts to get this draft to you well in advance of the date in deference to the scheduling difficulties of the end of the calendar year.

The guidelines for testimonials are at the bottom of this email. Deadlines for sending those documents to w3t-pr@w3.org is currently 7 January 2008.

Again congratulations on work well done. We look forward to having your participation in the Recommendation announcement.

Kind regards,

Janet Daly, W3C Communications Team

Guidelines for Testimonials

Supporting Testimonial Statements along W3C Press Releases

The Communications Team invites the Membership to participate in a press announcement by submitting supporting testimonial statements. Testimonials are not translated by W3C, but submission of testimonials in multiple languages is encouraged.

Please note that the press release itself is expected to be multilingual (at least available in English, French and Japanese).

Testimonial Guidelines

A testimonial must

contain a statement of support for the W3C and the technology we are promoting in this release;
be grammatically correct;
be attributed to an individual in the submitter's organization (name, title, org)

A testimonial should

contain an explanation of why the news is important to the Web or to specific communities, extolling the virtues and features not before enjoyed;
include a commitment for support of the spec in existing or future products and/or services

A testimonial may

include product announcements which support the technology now or plans to support later;
be in a language other than English.

A testimonial must not

contain criticisms of the technology that may undermine the success of the specification;
contain links to sites outside of W3C;
criticize other products or services;
exceed 100 words.

(note to self, this doesn't sound very "scientific"?)

Upon request by journalists, we would also pass on contact information of companies who have contributed testimonials.

Questions and Follow-up

If your marketing team has any questions, please have them phone Janet Daly (voice: +1 617.253.5884) or send mail to w3t-pr@w3.org.

shmontology

cariaso — Sun, 02 Dec 2007 14:15:11 -0500

big upfront design is suicide.

DBpedia has written rdfs for the existing wikipedia templates. I should have no problem creating rdfs for the snpedia ones. For the moment I'm keeping flexibility by writing all my queries in python. In time I'll carve the stable 80% into an rdf, and be able to use sparql.

From this web page you can play with wikipedia semantic queries.
http://wikipedia.aksw.org/index.php?qid=13

Mike Cariaso * Bioinformatics Software * http://www.cariaso.com

Dude, Where's Your Ontology?

Duncan — Sun, 02 Dec 2007 12:56:54 -0500

Hello Mike, had a look at SNPedia and the semantic wiki. Couldn't find the (sacred?) ontology (RDF / OWL / Whatever) anywhere, is it publicly available? Maybe I didn't look hard enough...

Anne Cregan at the University of New South Wales, Australia has kindly drawn my attention to an article in the December 2007 issue of Scientfic American "The Semantic Web In Action" which claims corporate applications of the semantic web are well under way, and consumer uses are emerging... haven't read it yet, but looks like it might be interesting.

Its written by Lee Feigenbaum (IBM), Ivan Herman (W3C), Tonya Hongsermeier , Eric Neumann and Susie Stephens .

warning believer

cariaso — Fri, 30 Nov 2007 12:21:28 -0500

A solution is a web with both, where some websites are more semantic than others. The semantic web lacks commercial success, because it has no place to put advertising. Best Buy has no motivation to make a semantic web site. But NCBI definitely does.

My own solution has been to ensure that data I produce is semantic friendly. The data is written onto wiki pages. The whole site becomes a semantic book about one topic. humans can edit it, but programs can crawl and query it. Humans and bots work together, on mix of wikipedia and relational database.

A public example is at
http://www.snpedia.com/index.php?title=SNPedia
which tracks human genetic variations and the tests for them

and these slides explaining a semantic wiki based LIMS
http://bioteam.net/wikilims.html

These are baby steps towards a full semantic web, but I've found them extremely powerful.

The algorithmic simplicity of ontologies solves many important problems, but fails miserably for much of the real world. Tag systems adapt better to the real world. Both can be modeled semantically, and should be.

Mike Cariaso * Bioinformatics Software * http://www.cariaso.com

GMOD at CSHL is also hiring

jason — Sat, 05 May 2007 15:11:46 -0400

Lincoln Stein's group is also hiring for GBrowse, BioMart, and Reactome development.

DK-heads

Duncan — Tue, 15 Aug 2006 09:33:13 -0400

Hi Jeremy, fancy meeting you here :) Its a shame DK have pulled the plug on this series, but these titles are all still available. Perhaps DK have trouble competing with Very Short Introductions and Icon Books "Introducing" series?

Hard / Soft sciences

Duncan — Tue, 15 Aug 2006 09:19:46 -0400

Hi Neil, I agree with you, the world would be a better place if more scientists were "intelligent people with wide-ranging interests who are able to grasp key concepts from outwith their own field". Unfortunately too many of them are experts in the own fields, and completely ignorant of others. All of the above seems fairly consistent with the Andy Brass rule that "Biology is the subject for people who like science but are scared of mathematics". IMHO you can run, but you can't hide from mathematics because it is fundamental to all the sciences. Biology and bioinformatics are no exception.

Thanks

JeremyCherfas — Mon, 14 Aug 2006 04:42:16 -0400

This is embarrasing. I came here (after picking up nodalpoint on my referrer logs) and read this article's first screenful. Decided it would be useful for a friend involved in managing a bioinformatics project, and wacked the link off to him. Then I read on. Now I will have to explain to my friend that I wasn't just doing a bit of ego-boosting. He'll understand, I hope. But I did want to thank Duncan for the good words.

The mystery, as far as I am concerned, is why Dorling Kindersley pulled the plug on what they promised us would be a long-running series that would grow. I know, the old joke about "how can you tell when a publisher is lying?" applies. But still, I did that book as a loss leader, on the promise of good promotion and more to come. I was fooled. But I remain proud of it and stand ready to help anyone else who wants to get complex ideas across.

perennial favourite

Neil — Sun, 13 Aug 2006 22:16:18 -0400

Ah yes, interdisciplinary communication, especially between biologists and programmers is a real favourite topic. Thanks for the article and links - it's good to see this addressed from the computer -> biology direction.

I'll start with my own story. A few years ago, I was given the task of redesigning a website for the group in which I worked. The original version had been created using some Mac WYSIWYG package. The boss came in one day to see how it was going and I pulled up some HTML onto the screen. He looked at it blankly. "What's this?" he said. "This isn't how I did it." "Um, I think it is", I replied, "this is HTML - it's what makes web pages appear." Eventually I realised that there was absolutely no connection in his mind between the HTML and the web page. Although he'd created a web page himself, his software had shielded him to such an extent that he had no concept of the link between code and output. So point one is this:

* Many biologists are passive users of computers. They are no more sophisticated than the average home user - they use Office, surf the web and read/send email. They do not even realise that to program means to make a computer do what you want or that in principle, with sufficient programming skill, anything is achievable.

This is something that has always perplexed me, because these same people are well aware of their limitations. They will spend hours doing tedious copy/paste operations which could be done in seconds using a small perl script. They will even say to you "there has to be a better way to do this". Yet they seem unwilling or incapable of taking the next step - teaching themselves some basic programming skills. The attitude is almost "but I'm a biologist, I just don't do that sort of thing".

In no particular order, here are a few other issues that I think are perpetuating the division between computational and non-computational biologists:

* The education system. We are still splitting people into "hard science" and "soft science" streams at too early an age (high school or early undergrad). It's still the case that people do maths, physics and computing because they care little for fluffy bunny rabbits and people do biology because they think maths, physics and computing is hard and irrelevant to life science. I think the only way to fix this is compulsory maths, physics and especially computing in all first year biology courses. I also think undergraduates need to be exposed to real research much earlier in their courses, so as they can see why these tools are useful.

* Microsoft. As many biologists are essentially home computer users, they believe the pointy-clicky GUI world of Windows to be "normal" and Linux/UNIX to be something niche and strange. In fact the reverse is true - UNIX has been around since 1970, Linux since 1990. This attitude seems to run through our entire society, as when PCs are advertised that can "run Windows as well as Mac", as though that's all there is. You'd think intelligent scientists would see past this - but no.

* IT support for biologists. If you work in an academic setting, IT support for biologists is likely to be woeful. Simply because as outlined above, your average biologist works like a home computer user and uses Mac or Windows. Campus IT services are therefore tailored for this - most of them don't even have a Linux expert on hand. If you want to setup a Linux server or build a cluster, you'll be on your own.

* The amateur nature of bioinformatics. The vast majority of practising bioinformaticians/computational biologists are people who started out as wet-lab research biologists and taught themselves computer programming, just out of interest. This creates a situation where we are seen as being in our own camp. Professional computer scientists take issue with our coding skills, biologists take issue with the fact that we've left the lab. Ideally we'd be at the interface where we can facilitate communication and research in both directions and this does happen in the right environment. Perhaps this will improve as more graduates with degrees in bioinformatics start to trickle out of the system.

I've always been of the opinion that anyone in science should just call themselves "a scientist". A scientist to me is an intelligent person with wide-ranging interests who is able to grasp key concepts from outwith their own field, even if they don't fully comprehend all of the details at first. I also think that we are living in incredibly exciting times. For me, the social changes in communication brought about by the internet and the transformation of biology from a soft science by genomics and computation are the key to blowing away boundaries between traditional research fields and creating truly interdisciplinary science. Unfortunately it's hard not to conclude that the way academia is set up just now is a substantial barrier to this ideal.