Guang Song receives NSF CAREER Award

March 8, 2010

Dr. Guang Song, Assistant Professor of Computer Science and faculty member of the Baker Center for Bioinformatics and Biological Statistics and the Bioinformatics and Computational Biology (BCB) graduate Program, has received an NSF CAREER award for his project on building a computational framework for mapping ligand migration channel networks and predicting molecular control mechanisms. 

The NSF Faculty Early Career Development (CAREER) Program is a Foundation-wide activity that offers the NSF's most prestigious awards in support of junior faculty who exemplify the role of teacher-scholars through outstanding research, excellent education and the integration of education and research within the context of the mission of their organizations. 

Project Title: CAREER: A Computational Framework for Mapping Ligand Migration Channel Networks and Predicting Molecular Control Mechanisms 

Project Description: 

Proteins are fundamental elements of living organisms. They are marvelous microscopic bio-machines that function in a steady, predictable manner. Together with other elements such as DNA, they make up the basics that underlie the complexity of life. The quest to know how these bio-machines work has inspired intense scientific curiosity and imagination. Since most functions are carried out dynamically and are difficult to observe directly from experiments, computational methods have an important, irreplaceable role to play. 

The objective of this project is to map out the ligand migration channel networks inside proteins and determine the molecular control mechanisms by which these channels are regulated dynamically. To overcome the limitations that existing methods face, the project will develop and employ a novel, efficient computational framework that draws one of its inspirations from path planning in robotics, as a ligand's migration in a dynamic protein resembles closely a mobile robot's navigation in a dynamic environment. The proposed approach will overcome the computational barrier by integrating efficient geometric mappings with the dynamic exploration of a protein's structure flexibility. By taking as input the structure ensemble of a given protein, which may be composed of existing experimental structures and/or conformations generated from molecular dynamics simulations, the proposed method carries out a spatial mapping of the protein's inner space at each conformation in the ensemble. The spatial mapping reveals the partial connectivity of the cavities and channels inside the protein. All partial maps are then merged to form a super-graph that represents the complete migration channel network that is accessible to the ligand, spatially and dynamically. The method will be applied to map out the ligand migration channel networks in a family of proteins and to study the similarities and differences in the ligand channel networks across the family, which is novel. Moreover, since the channel network is mapped for each individual conformation in the ensemble, direct correlation data between conformation changes and variations in channel sizes will be collected and analyzed to identify the key conformation changes that regulate these channels. Key residues that are responsible for the key conformation changes will then be identified.