Journal of Proteomics & Bioinformatics
Open Access

Interactomics: Computational analysis of novel drug opportunities

International Conference on Protein Engineering

October 26-28, 2015 Chicago, USA

Ram Samudrala

State University of New York, USA

Keynote: J Proteomics Bioinform

Abstract :

We have developed a Computational Analysis of Novel Drug Opportunities (CANDO) platform funded by a 2010 NIH Directorâ�?�?s Pioneer Award that analyses compound-proteome interaction signatures to determine drug behavior, in contrast to traditional single target approaches. The platform uses similarity of interaction signatures across all proteins as indicative of similar functional behavior and non-similar signatures (or regions of signatures) as indicative of off- and antitarget (side) effects, in effect inferring homology of compound/drug behavior at a proteomic level. We have created a matrix of predicted interactions between 3,733 human ingestible compounds including FDA approved drugs and supplements�?�?48,278 proteins using our hierarchical chem and bio-informatic fragment-based docking with dynamics protocol (from over one billion predicted interactions total). We applied our compound-proteome signature comparison and ranking approach to 2030 indications with one approved compound and yielded benchmarking accuracies of 12-25% for 1439 indications with more than approved compound. We are prospectively validating â�?�?high valueâ�? predictions in vitro, in vivo and by clinical studies for more than forty indications including dental caries, dengue, tuberculosis, ovarian cancer, cholangiocarinomas, among many fothers. 58/163 (36%) predictions over twelve studies across ten indications show comparable or better activity to existing therapies or micromolar inhibition at the cellular level and serve as novel repurposeable therapies. Our approach is applicable to any compound beyond those approved by the FDA and also include can readily consider mutations in protein structures to enable personalization based on genotype, foreshadowing a new era of faster, safer, better and cheaper drug discovery.

Biography :

Ram Samudrala is Professor and Chief, Division of Bioinformatics, State University of New York, Buffalo researching multi scale modeling of atomic, molecular, cellular and physiological systems with an emphasis on protein and proteome structure, function, interaction, design and evolution. His work has led to more than 115 publications in journals such as Science, Nature, PLoS Biology, the Proceedings of the National Academy of Sciences and the Journal of the American Medical Association. He has joined the University of Washington Faculty in 2001 (where he remains as an Affiliate Professor) after completing his Doctoral research with John Moult at the Center for Advanced Research in Biotechnology in 1997 and his Postdoctoral research with Michael Levitt (2013 Nobel in Chemistry) at Stanford University in 2000, which resulted in him making some of the best predictions at the first three community-wide assessment of protein structure prediction (CASP) experiments.

Email: ram@compbio.org