Kinase proteins, which regulate the activity of other proteins, are a major class of cancer therapy targets, with over 65 FDA-approved drugs targeted against them. However, tumors can evolve resistance to kinase-targeting therapies, and it remains difficult to predict whether a specific tumor will resist a particular kinase-targeting drug. Dr. Singh will use protein structural models and biophysical predictions to analyze how kinase mutations cause cancers to resist therapy. As these computationally intensive calculations could require decades on a single desktop computer, he will use a computing platform called Folding@home, which harnesses idle computer time donated by citizen scientists around the world to run the calculations. By developing new algorithms to predict whether a known mutation will resist a kinase-targeting drug, Dr. Singh hopes to advance precision oncology to allow clinicians to predict a treatment's chance of success given a patient's tumor profile. While his work primarily focuses on resistance to the drug crizotinib, used to treat non-small-cell lung carcinomas, his approaches can be extrapolated to other tumors and cancer targets. Dr. Singh received his BA and his PhD in computational and molecular biophysics from Washington University in St. Louis.
Molecular dynamics (MD) simulations are computational microscopes that model and capture atomically detailed protein motions. To analyze MD simulations, Dr. Singh will construct Markov State Models, network representations of a protein's conformational landscape, and couple them with information theoretic measures of communication between mutated residues and drug binding sites. Alchemical Free Energy calculations will predict the impact of mutation on a drug's binding energy using artificial "alchemical" intermediates to measure the energetic cost of mutating a residue.
