Скачать или смотреть The Biological, Algorithmic and Computational Challenges of Systems Biology, Rick Stevens

Breakthroughs in biology are being powered by advanced computing capabilities that enable researchers to manipu-late, explore and compare massive datasets. The speed at which any given scientific domain advances will soon depend on how well its researchers collaborate with one another, and with computer scientists and mathematicians. The ability to rapidly analyze new data, relate it to existing knowledge and derive new predictions to inform the next step is the key to accelerating scientific discovery.
In this talk I will discuss three things. First I’ll provide an overview of the new approach to biological research known as “systems biology” that addresses the problem of reverse engineering complex biological systems and using that in-formation to build integrative models of cells, pathways and networks and how we use these models to say things about the systems behavior of organisms and communities. I’ll then talk about a specific project to build a “systems knowl-edge base” known as the “KBase” project for the Depart-ment of Energy that is aimed at advancing predictive sys-tems biology in microbes, microbial communities and plants. The KBase project is integrating data from many existing sources, building tools and services that will support com-plex workflows enabling modeling of microbes, reconciling experimental data with computational predictions, and pro-viding a large-number of computational services that go be-yond existing integrated biological databases. KBase will be deployed on a purpose-built infrastructure spanning four laboratories that collectively house multiple petabytes of data, and that will support scalable computing resources on both cloud and cluster environments. End users will be able to access many thousands of public genomes and re-lated datasets for microbes. They will also gain access to tens of thousands of metagenomic samples and dozens of plant genomes and phenotype datasets. In addition to pro-viding web and programmatic interfaces to these data, the KBase will enable users to upload their own private data and virtually integrate it with the public datasets for com-parative analysis and development of models. The KBase is aiming to enable collaborative workflows and multiple ways of sharing. Finally, I’ll dive down into a couple of interesting examples addressing novel data organization and new algo-rithms that will be required to more effectively analyze the large volumes of data that are being assembled to support genome scale biological research. I’ll touch on alignment-free methods for sequence comparison, computing on compressed representations of closely related genomes and the problem of representing genome assemblies as graphs rather than lin-ear sequences.