Computational and Systems BiologyRecent collaborations have led to the following results:
Transcription in yeast is a noisy process With the Bertrand Lab, Montpellier, I have developed stochastic models to explain the cell-to-cell variability observed in single-cell mRNA expression data. Aitken et al. (2010) PLoS ONE 5(1): e8845.
Splicing has multiple rate-limiting steps A real-time single-cell imaging technique developed in the Bertrand Lab combined with systems modelling shows there to be four rate-limiting steps in co-transcriptional splicing. Schmidt et al. (2011) Journal of Cell Biology 193(5): 819.
Co-transcriptional splicing is more efficient In collaboration with the Beggs Lab, the kinetics of transcription and splicing in yeast have been modelled in detail. Splicing occurs while the RNA is being transcribed, and we have shown this pathway to be more efficient than post-transcriptional splicing. Aitken et al. (2011) PLoS Computational Biology 7(10): e1002215.
Prokaryotic transcription factor binding sites are often palindromic Under the supervision of Bruce Ward and myself, Alastair Kilpatrick is investigating improved heuristics for use in probabilistic models of transcription factor binding sites in prokaryotes. A measure of motif palindromicity is described in this poster: Kilpatrick et al. (2011) Rocky 9
InterPro signatures are predictive of enzymatic function Luna De Ferrari has recently published the EnzML technique for assigning enzymatic function to the proteins using their InterPro signatures as a feature space. EnzML can re-annotate entire proteomes with subset accuracy ranging from 87% for A. thaliana and reaching 97% for E. coli. De Ferrari et al. (2012) BMC Bioinformatics 13:61.
New methods for model comparison and parameter inference are being developed in a BBSRC strategic tools and resources project. Nested sampling will be used to compute both the Bayesian evidence and the bounds on optimal parameter values for systems models. More details can be found on the project pages.