LSB on Tuesday, August 15th

Session 1 (09:00‑10:00)

09:00‑09:50

Anne Condon (University of British Columbia) DNA and RNA Molecules: Glimpses Through an Algorithmic Lens RNA molecules are increasingly in the spotlight, in recognition of the important roles they are now known to play in our cells and their promise in therapeutics. Function follows form in the molecular world, and so our ability to understand RNA function is enhanced by reliable means for predicting RNA structure. Outside of the cell, exotic DNA structures are now finding use in the construction of biosensors, nanotubes, lattices, and much more, motivating the need for DNA structure prediction. Not surprisingly, computer scientists are interested in these structure prediction challenges -- and view RNA and DNA molecules in their own twisted way. In this talk, we'll describe successes and challenges of computational DNA and RNA structure prediction.

Break (10:00‑10:30)

Session 2 (10:30‑12:30)

10:30‑11:30

Larry Ruzzo (University of Washington) Searching for Noncoding RNA [pdf] Recent surprising experimental discoveries have revealed complex and unexpected roles for RNA molecules in most living organisms, beyond the well-known role of coding for proteins. Computational tools for the discovery of protein and DNA sequence elements have been actively studied for decades, but RNA molecules evolve in different ways, making the problems of discovering and searching for them very challenging. I will outline computational tools we have developed for these problems, both for discovery of RNA motifs from unaligned sequences and accelerated search for additional instances thereof, and describe preliminary results from our systematic searches for novel RNA elements in the genomes of hundreds of bacteria. Thousands of CPU hours (not to mention grad-student hours) have resulted in many promising candidates, some of which have already been verified experimentally.

11:30‑12:15

Rajeev Alur (University of Pennsylvania) Hybrid Systems Modeling for Regulatory Pathways [ppt]

Lunch break (12:30‑14:00)

Session 3 (14:00‑15:30)

14:00‑14:45

Ilya Shmulevich (Institute for Systems Biology) Insights from Boolean Modeling of Genetic Regulatory Networks [ppt] Complex dynamical networks of molecular interactions govern most of the biological processes and functions of a living cell. Mathematical and computational models of such networks form the basis of two main lines of inquiry. First is the broad goal of uncovering the underlying mechanisms and structure of genetic networks by making statistical inferences from large-scale measurements of biological systems under a variety of conditions. Concomitant with this goal is the possibility of using the inferred models to make predictions of how a cell will respond to specific perturbations. Second is the use of such models for gaining insight into the role of the topological and dynamical properties of such networks in governing the integrated behavior of the cell. Of particular interest is the manner in which a cell is able to respond to a diverse set of environmental conditions, while maintaining stability and robustness. I will discuss several approaches for addressing these questions using the modeling framework of Boolean networks and their probabilistic generalizations.

14:45‑15:30

Ashish Tiwari (SRI International) Symbolic Systems Biology: Hybrid Modeling and Analysis of Biological Networks [pdf] How do living cells compute and control themselves, and communicate with their environment? Modeling of biological systems is a first step towards analyzing such systems to answer that question. These systems often work at multiple time scales. We argue that hybrid systems offer a powerful mathematical formalism for modeling complex biological systems. Once a hybrid model is obtained, various analysis techniques developed for these systems can be applied to help uncover structure in the dynamics of biological systems of interest.

Break (15:30‑16:00)

Session 4 (16:00‑17:45)

16:00‑17:00

Carolyn Talcott (SRI International) Pathway Logic: A Logical Approach to Modeling Cellular Processes [pdf] Pathway Logic is an approach to modeling cellular processes based on symbolic logic. It allows one to model aspects of the structure and state of interacting components, to represent individual process steps (reactions) and to study possible ways a system could evolve using techniques based on logical inference. Reactions can be modeled at many levels of detail ranging from micro steps representing events such as phosphorlyation at specific sites or binding of protein domains to macro steps such as the results of signaling or metabolic modules. Given a network of reactions and a specification of cellular components one can use logical inference to query the network about possible reaction pathways and outcomes. In this talk we will explain how signal transduction reactions are represented in Pathway Logic, and show how these models can be visualized and queried using the Pathway Logic Assistant.

17:00‑17:45

William Noble (University of Washington) Two Bioinformatics Applications of Dynamic Bayesian Networks [ppt] In this talk, I will describe two quite different uses of Bayesian models. In the first, a simple hidden Markov model is applied to genomic data, in which position along the chromosome serves as the "time" dimension. The model is used in an unsupervised fashion to segment various genomic data sets, thereby revealing higher-order domains. We show that we obtain better fits using a model with a hierarchical emission distribution and with variable length duration modeling. The final segmentation yields insights into structural features of the human genome. In the second application, we investigate proteomic data from mass spectrometry. In this domain, the "time" dimension corresponds to locations along a short sequence of amino acids. The model captures statistical properties of the fragmentation of these sequences inside the mass spectrometer, and the model allows us to more accurately identify the amino acid sequence corresponding to an observed mass spectrum.