The goals of this project are to elucidate the fundamental principles underlying protein biochemical function and to apply the resulting insights on the interrelationships of protein structure, function and evolution to functional annotation and to assist in drug discovery. The underlying theme is that many features of protein structure and function arise from their physical properties without selection for function, which evolution then acts to fine tune/optimize. This will be demonstrated by the coincidence of the structural and functional properties of native proteins with artificially generated, homopolypeptide "SYN" protein structures to which protein-like sequences are added based on their stability in a particular fold. We will focus on ligand binding pockets and will explore the relationship between global fold similarity, pocket location/shape and amino acid conservation. A key objective is to demonstrate that ligand binding pocket geometry and sequence can be uncoupled from a protein's global structure, so that functional inference can be made between structurally different proteins. These ideas will extend our FINDSITEcomb Ligand Homology Modeling algorithm by developing approaches that better identify common ligand binding pockets and ligands in proteins having either similar or unrelated global folds. FINDSITEcomb will provide predicted structures, GO function, ligand binding sites, and virtual ligand screening/binding pose predictions. An important application will be to predict off-targets of FDA approved drugs in the human proteome. These off-target predictions will be experimentally tested for a significant number of proteins. To achieve these objectives, four Specific Aims are proposed: 1. Examination of the roles of physics and evolution in determining protein structure and function. 2. Exploiting insights from Aim 1, a new approach to difficult target threading will be developed. 3. The major limitations of FINDSITEcomb will be addressed and the methodology applied to drug repurposing. 4. Using thermal shift assays, FINDSITEcomb's virtual screening predictions will be experimentally tested. The entire computational methodology will be applied to the human proteome and model organisms, and the SUNPRO database will report all results. All tools will be made available on our superPSIFR webserver and as downloadable software, including source code.
These aims represent an integrated effort to elucidate the principles underlying protein structure and function and to apply these insights to improve function predictions.

Public Health Relevance

By providing insights into the biochemical functions of the plethora of unannotated proteins provided by the genome sequencing efforts, computational approaches can help address the increasing dichotomy between having a protein's sequence and knowledge of what it does. The proposed research will exploit fundamental insights into biological function to develop new automated structure-based algorithms for the prediction of protein structure and function. By predicting new protein targets of FDA approved drugs, which are then experimentally validated, it could help develop new therapeutic approaches for the treatment of diseases.

Agency
National Institute of Health (NIH)
Type
Research Project (R01)
Project #
5R01GM048835-21
Application #
8723239
Study Section
Macromolecular Structure and Function D Study Section (MSFD)
Program Officer
Wehrle, Janna P
Project Start
Project End
Budget Start
Budget End
Support Year
21
Fiscal Year
2014
Total Cost
Indirect Cost
Name
Georgia Institute of Technology
Department
Biology
Type
Schools of Arts and Sciences
DUNS #
City
Atlanta
State
GA
Country
United States
Zip Code
30332
Zhou, Hongyi; Skolnick, Jeffrey (2016) A knowledge-based approach for predicting gene-disease associations. Bioinformatics 32:2831-8
Skolnick, Jeffrey; Gao, Mu; Zhou, Hongyi (2016) How special is the biochemical function of native proteins? F1000Res 5:
Srinivasan, Bharath; Marks, Hanna; Mitra, Sreyoshi et al. (2016) Catalytic and substrate promiscuity: distinct multiple chemistries catalysed by the phosphatase domain of receptor protein tyrosine phosphatase. Biochem J 473:2165-77
Srinivasan, Bharath; Skolnick, Jeffrey (2015) Insights into the slow-onset tight-binding inhibition of Escherichia coli dihydrofolate reductase: detailed mechanistic characterization of pyrrolo [3,2-f] quinazoline-1,3-diamine and its derivatives as novel tight-binding inhibitors. FEBS J 282:1922-38
Zhou, Hongyi; Gao, Mu; Skolnick, Jeffrey (2015) Comprehensive prediction of drug-protein interactions and side effects for the human proteome. Sci Rep 5:11090
Lee, Hui Sun; Jo, Sunhwan; Mukherjee, Srayanta et al. (2015) GS-align for glycan structure alignment and similarity measurement. Bioinformatics 31:2653-9
Gao, Mu; Zhou, Hongyi; Skolnick, Jeffrey (2015) Insights into Disease-Associated Mutations in the Human Proteome through Protein Structural Analysis. Structure 23:1362-9
Skolnick, Jeffrey; Gao, Mu; Roy, Ambrish et al. (2015) Implications of the small number of distinct ligand binding pockets in proteins for drug discovery, evolution and biochemical function. Bioorg Med Chem Lett 25:1163-70
Roy, Ambrish; Srinivasan, Bharath; Skolnick, Jeffrey (2015) PoLi: A Virtual Screening Pipeline Based on Template Pocket and Ligand Similarity. J Chem Inf Model 55:1757-70
Roy, Ambrish; Skolnick, Jeffrey (2015) LIGSIFT: an open-source tool for ligand structural alignment and virtual screening. Bioinformatics 31:539-44

Showing the most recent 10 out of 114 publications