BBSRC Systems Biology Grantholder Workshop, University of Nottingham, 16 December 2008.
David Wild and Global Metabolic Switching in S.coelicolor
a.k.a STREAM, or Streptomycete analysis of metabolism. The aim of this project is to understand switching between primary and secondary metabolism. I didn't take many notes as my computer's battery was beginning to run down so I gave it a break (sorry – but there are no plugs in the auditorium!). Then there was the joint talk of Stephen Muggleton and Emily Kay from CISBIC, and the experiments and pathway modelling of polysaccharide structures from Campylobacter jejuni.
Stephen Muggleton: Among other things, they have adaptated the Robot Scientist project for this project. They are looking at host-pathogen interactions. They have also experimented with various machine learning techniques, including probabilistic logic programming (PLP) or inductive LP (ILP). PLP/ILP takes background knowledge (e.g. protein seq, partial grammar, incomplete biological networks) together with examples. These inputs allow the generation of hypotheses. CISBIC Sub-project 1 involves two species: C.jejuni and M.bovis. Focussed on synthetic pathways related to construction of capsule structure for C.jejuni: 38 genes involved that are fully sequenced, with a complete set of knock-outs. Only 2/3 of the functions are known or suspected (which makes it harder for the machine learning side of things). They started by building up a prolog database of know information about C.jejuni (KEGG, BioCyc, Brendan & other papers). From this database, they perform the machine learning and visualization. They use an abductive framework (a technique for generating hypotheses) for modelling the genotype-phenotype relations, mainly for gap-filling in networks. Predictions include Cj1432, a protein of unknown function, is central to capsule pathway (this is not something that can be determined by BLAST and amino acid similarity). The initial tests on Cj1432 do suggest the absence of capsule, so looks good, but more stringent tests need to be performed.
Emily Kay: (again, I didn't get all of her talk because my battery finally gave way. Also, her talk was much more wet-lab oriented, which I have a harder time following. My apologies! ) C.jejuni is gram negative and quite difficult to cultivate in the lab. Usually present as a commensal in avians. It's the leading cause of bacterial diarrhoeal disease worldwide. Transmission is often through consumption or handling of contaminated poultry products. Associated neuropathies include Guillain-Barre syndrome, which is a form of paralysis. Some questions: why are some strains pathogenic and others aren't? ; Why are they commensal in avians and highly infectious in humans? Answers likely to lie in glycostructures on the surface. However, how do we get from sequence to sugar nucleotides? How are these regulated? There are a number of Campylobacter resources: small tractable genome, significant strain variation to infer biologically relevant questions, whole-genome data from 11 jejuni strains, access to pan-species jejuni microarrays, etc. Capsule probably helps against dessication and gives it the ability to live outside the host, at least for a time. Data that can be used for modelling include: microarray data, metabolomic data, ppi data, and biochemical network for glycosylation.
These are just my notes and are not guaranteed to be correct. Please feel free to let me know about any errors, which are all my fault and not the fault of the speaker. 🙂