NSF Bioinformatics Grants Awarded Sept. 26 — Dec. 17, 2009
Multiscale Stochastic Simulation for Complex Biochemical Systems with Visualization Tools
Start date: June 1, 2010
Expires: May 31, 2011
Awarded amount to date: $98,020
Principal investigator: Yang Cao
Sponsor: Virginia Polytechnic Institute and State University
This project aims to develop efficient simulation techniques for stochastic biochemical models, particularly the cell cycle model. "Biologists have developed complex mathematical models of cell-cycle control in budding yeast, fission yeast, and mammalian cells," the grant abstract states. "These systems are so complex that its simulation and analysis present great challenges to computational science." In response, the investigators plan to develop "innovative computational methods and rigorous mathematical theories to integrate the full gamut of continuous, discrete, deterministic, and stochastic models, and support dynamic, seamless and automatic switching between different models and algorithms as dictated by the scales of underlying problems." The primary aim is to develop algorithms and mathematical theories about a particular multiscale challenge known as "stiffness." In particular, they plan to "develop the theory of the stiffness in discrete stochastic simulation of chemically reacting systems and an automatic stiffness detection algorithm through a running-time profile analysis." The team also plans to develop hybrid algorithms to simulate biological systems with multistate species, "a special challenge in biological systems with multiple binding sites."
Robustness in Genetic Regulatory Network Modeling and Control
Start date: Feb. 1, 2010
Expires: Jan. 31, 2011
Awarded amount to date: $77,107
Principal investigator: Ranadip Pal
Sponsor: Texas Tech University
The goal of this project is to develop a theoretical and computational framework "to estimate the uncertainties in genetic regulatory network modeling and generate robust therapeutic strategies for genetic diseases such as cancer," according to the grant abstract. The investigators note that one role for genetic regulatory network modeling is "to design and analyze therapeutic intervention strategies aimed at moving the network out of undesirable states, such as those associated with disease, and into desirable ones." However, they add that limited experimental data "prevent accurate inference of the mathematical model of the genetic regulatory network." In response, the team plans to study the effect of modeling errors on the predictive power of inferred network models and on intervention outcomes. They also plan to design control strategies that will provide bounds on the performance of mathematically designed intervention strategies. In particular, they will develop algorithms to design robust control strategies "with the objectives of avoiding extremely undesirable results (minimax design) or improving the expected chances of success (Bayesian approach)."
Computational Models and Algorithms for Differential Network Analysis in Systems Biology
Start date: Feb. 1, 2010
Expires: Jan. 31, 2011
Awarded amount to date: $80,275
Principal investigator: Mehmet Koyuturk
Sponsor: Case Western Reserve University
This project aims to develop computational models and algorithms that enable integration of disparate datasets — including gene sequences, gene expression, protein expression, and protein interactions — for network-based modeling of phenotypic differences. The investigators plan to develop innovative methods to conduct combinatorial modeling of coordinate changes in the expression of multiple interacting genes; probabilistic modeling of the crosstalk in biomolecular networks; and algebraic modeling of information flow in the cell.
Molecular Dynamics Simulations of the Conformational Binding Dynamics of Sequence-Specific DNA-Binding Proteins
Start date: Oct. 1, 2009
Expires: Jan. 31, 2010
Awarded amount to date: $91,946
Principal investigator: Arjan Van Der Vaart
This award supports the development of computational methods for simulating conformational binding dynamics of sequence-specific DNA-binding proteins. The project will focus on the lac repressor headpiece, which partially folds upon DNA binding, and the Ets-1 transcription factor, which partially unfolds upon DNA binding. "New analysis methods and biasing algorithms are being developed along with a coarse-grained model to compute transition pathways," according to the grant abstract.
Data-Parallel Platform for Large-Scale Simulation of Agent-Based Models in Systems Biology
Start date: Aug. 16, 2009
Expires: July 31, 2012
Awarded amount to date: $349,910
Principal investigator: Roshan D'souza
Sponsor: University of Wisconsin-Milwaukee
This award, funded under the American Recovery and Reinvestment Act of 2009, supports research into algorithms to model multi-scale phenomena in biological systems using agent-based models. It also supports the acquisition of two graphics processing unit racks. "GPUs follow the stream programming model," which is "completely different from traditional serial programming and requires the invention of entirely news set of algorithms to suit the architecture," the grant abstract states.