drinking the koolaid
PLoS Computational Biology has published a series of “Education” articles lately that are useful for those entering Bioinformatics / Computational Biology. Even seasoned veterans might find a nugget or two in these:
A Quick Guide for Developing Effective Bioinformatics Programming Skills
These joint RECOMB Satellite conferences were held at the Broad Institute (Oct 29 – Nov 2, 2008). This was a somewhat grueling 5 day meeting … with half hour keynotes (17) and twelve minute talks (>75). As with previous conferences, there is FriendFeed coverage. The full length papers from the conference will be published in either Genome Research, the Journal of Computational Biology, or Molecular Systems Biology … available by the end of the calendar year. As is frequently the case at computational biology meetings, some of the best talks were about recently published work.
My Highlights from the meeting:
Gal Chechik, Eugene Oh, Oliver Rando, Jonathan Weissman, Aviv Regev, Daphne Koller (2008). Activity motifs reveal principles of timing in transcriptional control of the yeast metabolic network Nature Biotechnology DOI: 10.1038/nbt.1499
Tomer Shlomi, Moran N Cabili, Markus J Herrgård, Bernhard Ø Palsson, Eytan Ruppin (2008). Network-based prediction of human tissue-specific metabolism Nature Biotechnology, 26 (9), 1003-1010 DOI: 10.1038/nbt.1487
M. Rabani, M. Kertesz, E. Segal (2008). Computational prediction of RNA structural motifs involved in posttranscriptional regulatory processes Proceedings of the National Academy of Sciences, 105 (39), 14885-14890 DOI: 10.1073/pnas.0803169105
Charles Kemp and Joshua Tenenbaum have published (or have in press) a series of beautiful papers concerned with using Bayesian inference. Their recent PNAS article opens with this sentence:
Discovering the underlying structure of a set of entities is a fundamental challenge for scientists and children alike. (Kemp PNAS)
The paper is focused on showing that discoveries about structural form can be understood computationally as probabilistic inferences about the organizing principles of a dataset.
The internal structure of the nucleus is complex. It is packed with assorted molecules all trying to work together to execute cellular function. The interplay of nuclear components such as transcription factors and nucleosomes is likely more complex than previously anticipated. A recent paper by Lam, Steger, and O’Shea beings to highlight that complexity. But first, the a little back story ….
A relatively recent read of Yu and Gerstein’s 2006 PNAS paper has got me thinking about analogies and metaphors in biological networks. As Yu points out, protein-protein interaction networks have frequently been compared to social communication networks … sharing characteristics such as small-world and scale-free properties. Yu presents a plausible analogy between transcriptional regulatory networks and organizational control structures (in other words, a form of bureaucracy). The difference, as Yu states is, “more oriented towards control than communication.”
I’ve been reading and thinking quite a bit lately about biological networks with respect to genetic interactions such as synthetic lethal (& synthetic sick), rescue, and dosage compensation. Oversimplifying, in all these cases the network is able to adjust to perturbation … sometimes in, as of yet, unpredictable fashion. The buzz words here are redundancy and robustness. A decent analogy already exist in this arena. Primarily associated with metabolic networks, the pipe analogy represents the pathways through the network as pipes which flow metabolites to reach some final destination. So a perturbation is effectively a disruption to a pipe. For robustness you depend on some other pipe in the system to pick up the load.
