Epidemiologic Methods II

Alexey Petrov, postdoctoral fellow (Joseph Puglisi Lab), Stanford University Multiple Parallel Pathways of Translation Initiation on Cricket Paralysis Virus Internal Ribosome Entry Sites http://bayviro.org/symposium/5/

Eric Schadt, Ph.D., Pacific Biosciences Environmental concerns over the use of fossil fuels and their role in climate change have sparked research on the development of alternative fuels. Hydrogen is a clean burning alternative fuel that can be produced in large amounts by some bacteria. We have embarked on a systems level approach to dissect metabolic and regulatory networks necessary for nitrogenase-catalyzed hydrogen production by the bacterium Rhodopseudomonas. Rhodopseudomonas is an ideal platform to develop as a biocatalyst because it is an extremely versatile microbe that produces copious amounts of hydrogen gas by drawing on abundant natural resources of sunlight and biomass. Hydrogen production requires the integration of dozens of metabolic reactions carried out in the context of a complex network of molecular interactions. To identify key drivers for hydrogen production we employed an integrative genomics approach similar to that used with great success to elucidate the causes of common human diseases. We have sequenced the complete genomes and transcriptomes of approximately 100 independently isolated Rhodopseudomonas strains. In contrast to mammalian systems, the genomes of these strains are quite different from each other and so a completely accurate genome assembly is required for each strain in order to leverage DNA variation as a systematic source of perturbations for network reconstruction. A breakthrough for this project is the first de novo assembly of bacterial genomes using reads from Pacific Biosciences' single-molecule real-time DNA sequencing platform using a novel assembly pipeline tailored to this type of sequencing. This combined with transcriptomic and phenotypic data from 100 strains grown under three different conditions enabled the construction of whole genome, causal gene networks using Bayesian network reconstruction methods, resulting in the identification of subnetworks supported as causal for hydrogen production and nitrogenase activity. Directed perturbations of these networks provide a path to enhance hydrogen production. http://ccb.berkeley.edu/

General Biology Lecture