February 21st CTE Seminar

Title:  Accurate and Efficient Gene Function Prediction using a Multi-Bacterial Network

Presenter: Jeffrey N. Law

Research Group: Computational Systems Biology

Abstract: The dramatic decrease in sequencing costs has led to hundreds of thousands of newly characterized prokaryotic genomes. Unfortunately, fewer than 0.01% of genes in these genomes have had their functions determined experimentally. Therefore, many computational methods to supplement experimental functional annotations have been developed. Despite these efforts, as many as 40% of genes in sequenced genomes lack any experimentally-determined or computationally-predicted function. To address this gap, we seek to develop methods for gene function prediction that integrate heterogeneous data for multiple species while also operating on a genomewide scale. However, the large size of such multi-species networks pose a challenge for the scalability of current state-of-the-art methods which typically operate on a single species or a small group of genes at a time. Inspired by this challenge, we develop a novel iterative label propagation algorithm called FastSinkSource. By using mathematically-provable bounds on the rate of progress of FastSinkSource to develop a new convergence strategy, we decrease the running time by a factor of 100 or more without sacrificing prediction accuracy. We systematically compare and evaluate many approaches to construct a multi-species bacterial network and apply FastSinkSource along with other state-of- the-art methods to these networks. We find that by pre-computing scores for species with experimentally-validated annotations and then transferring those scores to other species, FastSinkSource is able to make the most accurate functional predictions for 200 bacterial species, taking under 4 minutes for this computation. Our results point to the feasibility and promise of multi-species, genomewide gene function prediction, especially as more experimental data and annotations become available for a diverse variety of organisms.

Title: Modeling and analysis of nutrient signaling in Saccharomyces cerevisiae

Presenter: Amogh Jalihal

Research Group: Computational Systems Biology

Abstract: The coordination of metabolism with nutrient availability is essential for the survival and proliferation of all cells. In yeast, this coordination is achieved by dedicated nutrient sensing and signaling pathways that converge on key regulatory proteins that govern nutrient specific stress and adaptation responses, as well as general growth responses. We study the nutrient signaling system in budding yeast by creating a literature-curated hypothesis on the regulatory network. We transform this network into a system of differential equations, and calibrate the behavior of this dynamical model with data curated from the literature. We next carry out an extensive parameter robustness analysis in order to study the constraints imposed by the experimental data. Using an ensemble of parameter sets discovered as a result of the robustness analysis, we use the model to make phenotypic predictions in a variety of gene deletion mutants in a range of nutrient conditions that, if validated experimentally, can help refine model structure. The model of nutrient signaling that we have proposed is well poised to serve as the interface between the regulation of cellular metabolism and key decision making processes like the cell cycle, entry into meiosis, and autophagy.