And speaking of kickass jobs...

... NPG is hiring again.

Kickass job in the Holmes lab

Know AJAX? Remember that Google Maps style genome browser (no more reloading the page to scroll a little bit to the left)? Want to work on it?

The ideal applicant will have experience in two or more of the following areas:

• Javascript/AJAX programming; general web development
• Prototype, Dojo, Rico, script.aculo.us, Google Web Toolkit, AJAXSLT and/or other open-source Javascript libraries
• Cross-browser portability issues
• XML and XSLT
• Perl, Python, Ruby or other scripting language(s)
• SQL databases; My SQL or PostGreSQL
• Computational biology/bioinformatics

The post is based in the Holmes lab at UC Berkeley. More details can be found on biowiki.org.

EuroFoo

I'm in Brussels at the moment after attending EuroFoo over the weekend. As a format Foo and Bar camps rock: people talk about what they want when they want (well, in fixed timeslots, anyway) and the sessions are usually small enough for audience participation. I talked a little about Postgenomic and discovered that I suck at Pecha Kucha (tm) in the process. The valuable lesson learned was to always rehearse your presentation out loud at least once to get the timings down.

I had an interesting discussion about microformats vs "proper" semantic markup in the form of RDFa (see also the RDFa primer).

Here's a simple example of RDFa in action (from this presentation):

Normal link:

<a href="http://cc.org/licenses/by/2.5/">Creative Commons License.</a>

With RDFa:

<a xmlns:cc="http://cc.org/ns#" rel="cc:license" href="http://cc.org/licenses/by/2.5/"> Creative Commons License</a>

If you're into semantic markup and haven't already, check out the spec.

Anyway, the problem that RDFa has is that microformats are already out there (at Yahoo!, Techhnorati and lots of blogs) - and broadly speaking they work, despite their limitations. I think that it'll all come down to the tools: the first group that gets a WYSIWYG "add an address" or "add a review" button built into Wordpress, Moveable Type and Dreamweaver will win.

Without tools it seems to me like RDFa has a distinct disadvantage: it's easy to remember to add 'hCard' to a div to start marking up your contact details... would you remember to add xmlns:v="http://www.w3.org/2001/vcard-rdf/3.0#"?

Postgenomic / Pubmed mashup

Yet more Postgenomic stuff - pgcomments.user.js is a Greasemonkey script that adds blog post trackbacks to PubMed.

Here's a page to try it out on. (update: today this is giving XML related errors on the PubMed side - that's nothing to do with the script, really. :))

The script has a couple of known limitations:

• The trackbacks box will be positioned correctly if you're using the AbstractPlus view but otherwise it just gets added to the top of the page - which isn't ideal, but hey. :)
  
• Some papers are indexed by Postgenomic but don't have any posts associated with them (usually this is because they've been cited by Nature Highlights or the Faculty of 1000). In those cases the trackbacks box will appear but it'll be empty.
  
• The list of DOIs / PMIDs etc. in Postgenomic is around 161 kb in size. This means that sometimes there might be a delay before the trackbacks box appears - Greasemonkey has to download it in the background if an up to date version isn't already cached.
  

Let me know if you have any trouble.

Postgenomic / Connotea mashup

Inspired by Pedro's original Greasemonkey script I've written pgconnotea.user.js, which adds a comment icon (see the speech bubbles next to the DOI in the image above) next to Connotea bookmarks representing papers indexed by Postgenomic.

At the moment it works by checking DOIs - a great next step would be using the JSON results from the Postgenomic API instead (see the comments at the top of the script) to check for OAI / Pubmed identifiers too.

update: here's a link to a Connotea page that definitely contains a paper listed in Postgenomic. If you're trying the script out and you don't see any comment icons there, even after a refresh, drop me a line so that I can try and figure out what's wrong.

And the hottest scienceblogger is...

... at the end of the post. But first a disclaimer: to be included you had to (a) have a 'working science' blog indexed by Postgenomic, (b) have claimed your blog on Technorati and (c) have uploaded a portrait there. That's why the set of bloggers represented is relatively small.

Anyway, the results:

It's looking pretty bad for male sciencebloggers. People prefer the monkeys.

	Number	Average Hotness
Women	4	89.25
Primates (inc. those stuffed or crudely anthromorphised)	3	68
Other non-human portraits	21	53.38
Men	37	47.02

If you want some context, know that the 'skulls and skeletons' subcategory of 'other non-human portraits' has an average hotness of 56.

Is it a clean shaven thing? Chicks don't dig beards.

	Number	Average Hotness
Beardy Men	12	39.58
Non-Beardy Men	25	52.76

PZ Myers is a beardy exception, with a respectable hotness of 80.

I thought that it might be interesting to see how different types of scientist compare. Apparently climate change and medicine is hot, biotech is not.

	Number	Average Hotness
Earth and Environment	5	65.80
Neuroscience	9	64.4
Clinical Practice & Research	4	62.25
Science and Society	7	58.70
Physics	7	55.15
Bioinformatics	8	55.12
Natural History	3	54.33
Life Sciences	23	54.34
Biotechnology	4	36.00

And now what you've really come here for: who are the five hottest sciencebloggers on the planet (whose blogs are indexed by Postgenomic and claimed on Technorati by a user with a portrait which isn't of a stuffed monkey, a moose or a pair of underpants)?

1. Retrospectacle

2. Chaotic Utopia

3. The Scientific Activist

4. Pharyngula

5. Notes from the Biomass

Who's the hottest scienceblogger?

More playing around with the Postgenomic API.

Postgenomic collects a lot of stats. The relative hotness of different bloggers is not one of these stats. This must be rectified.

I'll put up an analysis of the results tommorrow.

voting has now closed

Changes

I started a new job recently at Nature Publishing Group, where I'm working with Alf Eaton in the web publishing department. Most of my time is going into developing a new kickass information management tool but I'll also keep working on Postgenomic (which remains open source: think Connotea).

While I enjoyed working in my previous lab it's refreshing to be in an environment where people are excited about the same ideas and opportunities as I am. I'm also not required to change printer cartridges any more (at least not yet), which is a plus.

On a side note (hur hur) I've been playing around with the new, long-time-in-coming Postgenomic API. Under "most popular posts" to the right you should see a list of Flags and Lollipops posts ordered by the number of incoming links. Nifty, no?

I refactored a lot of the Postgenomic code recently. At the time of writing postgenomic.com is still running an earlier version for technical reasons, but Sourceforge should have a new release soon. The new API gives you access to posts, blogs and papers ordered and filtered in various ways - metadata includes things like lat / long for conference reports, DOIs and pubmed IDs for papers and the number of incoming links (from other science blogs) for blogs and posts.

Things are still pretty alpha, but if any of that sounds interesting and you fancy getting your hands on an early release that's subject to change get in touch and I'll send you the relevant links.