Research Summary

Senior Fellowship Research Summary

Accurately modeling the 3D structures of proteins and their complexes is crucial to gain insights into the molecular mechanisms behind biological processes. This is especially true in systems where experimental information is scanty or difficult to obtain. Our modeling strategy consists of first studying the 3D structures of known complexes and extracting spatial restraints between different chemical groups/residues. Models of protein-protein, protein-ligand and protein-DNA complexes are then built by a satisfying these restraints. We intend on obtaining the necessary spatial restraints using structural comparisons done using our software CLICK ( CLICK recognizes structural similarity in structures and sub-structures of molecules, irrespective of their topology. CLICK would then be used to categorize different types of binding sites and these would in turn be used as templates to derive spatial restraints to model the geometry of interaction sites. Further, using this statistical knowledge we intend on designing/engineering protein-protein and protein-ligand complexes when the structure of only one of the interaction partners is known. We are hopeful that such computations would lead to accurate identification of drug targets and help in designing potent therapeutic agents and vaccines. 

A superimposition of 2 protein-DNA complexes (pdb codes 1ysa and 2ayg, shown as ribbons) using our software CLICK. Superimpositions such as this one, help determine the common factors in DNA recognition by diverse proteins. The 2 transcription factors shown in this example are not related in sequence or structure, yet the Arginine residues (shown in red in the alignment and in ball and stick representation in the figure) that mediate the protein-DNA interaction are in almost identical spatial locations. CLICK is a powerful computational tool to determine such similarities and could be used in functional annotation and molecular design.