S137: Functional annotation of microbial gene families: the evolutionary disconnect

Interest in microbial biofuel production has spurred genome sequencing projects for numerous bacteria that degrade plant biomass, archaea that produce hydrogen or methane, phototrophs that form lipids, and microbes that ferment carbohydrates to produce alcohols. In order to understand the metabolic systems of these microbes and engineer new functionality or robustness, genome sequences are expected to provide the scaffold for systems biology experimentation. Metabolic modeling and pathway optimization processes require high-quality gene function predictions that support automated pathway reconstruction and carbon and electron balance experiments. However, for reactions in many pathways key enzymes are unknown or missing. For other reactions, multiple isozymes are identified (or misidentified). Several major pitfalls in the prediction of enzyme function will be discussed, focusing on the convergent evolution of pathways and enzymes. We have identified evolutionarily unrelated pathways of sulfonic acid group formation that share a common oxygen-independent mechanism. We have also characterized substrate specificity determinants in 2-oxoacid elongation pathways for the design of next-generation bioproducts. Recent and upcoming advances in gene function prediction will be discussed, including strategies to combine structure-function experiments and molecular evolutionary analyses to enhance predictions.