Briefings in Bioinformatics: computational proteomics issue

Every so often, a new issue of the under-rated Briefings in Bioinformatics appears in my feed reader.

The latest is a special issue on computational proteomics. High-throughput proteomics is all the rage in academic, clinical and industrial settings just now, so this is well worth a read.

Bioinformaticians looking for ways to help out with the management and analysis of proteomic data should look in particular at:

Bioinformaticians in the service of bench biologists

Stumbled out of bed to the feed reader and came close to spraying cereal over the screen when I read this exchange on a Nature Network blog:

Original post:

Like them or loathe them, it’s not really possible to analyze a genome-wide screen without a large number of [Excel spreadsheets]

Comment #1 from our Pierre:

Oh please, please, please, no, don’t that with excel, please

He’s quite right, of course. Unfortunately, the ensuing debate is heading down a familiar track: “that’s all very well for you hardcore computer types, but we’re just simple bench biologists”.

Well look – a lot of us “computer types” were, or are, bench biologists too. We weren’t born with magical computer skills, nor did we learn them overnight. We know what we know and recommend it to others not out of geekiness or snobbery, but because we believe that if there’s a better way to perform a task, we owe it to ourselves to learn it. If others can’t make that commitment, we’re more than happy to help out and share what we’ve learned.

Just be prepared to meet us half-way, OK?

Evolution of an idea

It’s great to sit back and watch ideas and software unfold.

Just over a year ago, Euan asked whether anyone was employing AJAX in graphical genome browsers. The old-style “reload on refresh” browsers (UCSC, Gbrowse, Ensembl) were starting to look a bit Web 1.0.

This sparked plenty of discussion, including a pointer to X:Map: a very nice alternative view of Ensembl data using the Google Maps API (update: and of course ajax-ification of Gbrowse).

Jump forward to today and thanks to Euan’s feed via FriendFeed, I discover Genome Projector, which takes the zoom-able Google Maps idea to a new level.

And that’s how social networks let you discover stuff. Brilliant.


It’s online, so I guess I can tell you about:

Lonic, A. Barry, E.F., Quach, C., Kobe, B., Saunders, N.F.W. and Guthridge, M.A. (2008)
FGFR2 phosphorylation on Serine 779 couples to 14-3-3 and regulates cell survival and proliferation.
Mol. Cell. Biol. (ahead of print); DOI:10.1128/MCB.01837-07 [Abstract] | [Manuscript]

A minor contribution from me: they asked which kinases might phosphorylate S779, I gave them a list (using a tool that may see the light of day eventually), they showed that activation of a candidate kinase leads to increased phosphorylation. That would rate an acknowledgement from some people, but these guys were kind enough to add our names to the paper.

Just another scene from the life of the “go-to” bioinformatician.

Rewards, output and academia

Academia takes a very narrow view of what constitutes “output”. Rewards (such as funding or tenure) are given out on the basis of (1) publications, preferably first-author, preferably in so-called high-impact journals; (2) citations, in the same journals and (3) previous rewards – “demonstrated ability in securing funding”. I always find that last catch-22 clause particularly amusing.

I started to think about this when I read What is principal component analysis? [DOI 10.1038/nbt0308-303], in the current issue of Nature Biotechnology (subscription only). Now, I’m not criticising the article or its publication: it’s well-written, educational and a good basic overview of PCA for biologists who have not previously encountered the method. However, my first reaction was to recall a number of excellent blog posts on the same topic that I’ve read recently. For example:

The Nature Biotechnology article is recognised by academia and qualifies for academic rewards. The blog posts – which are longer, more detailed, written by enthusiastic communicators and in theory, accessible to a much wider audience (as opposed to people with a subscription to Nature Biotechnology) – are not.

It doesn’t seem right to me. How does your institution evaluate and reward “non-traditional” output?


In yet another moment of BBGM synchronicity, I started to think about lifestreaming and its applications as Deepak wrote about it. My inspiration was the recent article 35 ways to stream your life.

I’ve tried (and you can find me at):

  • Mugshot – aggregates a limited number of sources, doesn’t seem to update properly from, has conversation features (quips, comments)
  • FriendFeed – nice look and feel, a limited number of sources, has conversation features (comments, ratings)
  • Profilactic – by far my favourite in terms of look/feel and sources (you can add anything that has a feed) but no conversations as yet

Lifestreams are fun. I don’t expect anyone to care about what I just played on (and likewise), but these are all ways of broadcasting yourself and making connections. Read Deepak’s post for some thoughts on how this might apply to science.

Here’s a crazy idea – the workstream:

  • Neil parsed SwissProt entry Q38897 using parser script
  • Bob calculated all intersubunit contacts in PDB entry 2jdq using CCP4 package contact