Identifying Mutational Hotspots using Differences in Atomic Fluctuations combined with Positional Variability

Abstract

ABSTRACTAmino acid substitutions in proteins resulting from gene mutations are essential to the evolutionary process that enhances the organism’s adaption and survival rate. The primary aim of such substitution is to improve the protein’s function and structural stability. Various computational and experimental techniques are available to understand the effect of amino acid substitution on a protein’s function and structural stability and ultimately the effect on drug binding. This work seeks to understand how amino acid substitutions help overcome the anti-infective effect of drugs by diminishing their inhibitory potential. We attempt to locate hotspots among the amino acids in the binding site, where substitutions are likely to cause drug resistance. To achieve this, we propose an empirical score using atomic fluctuations coupled with amino acid variability to identify amino acids that are likely to mutate, leading to drug resistance. The scoring method has been developed and validated on two extensively studied drugs inhibiting HIV-1 protease, namely nelfinavir and darunavir. Further, we choose anaplastic lymphoma kinase (ALK) and its inhibitor, crizotinib, to test the performance of this empirical score. The prediction works well for amino acids within a radius of 5 Å from the centre of mass of the drug. The advantage of this method is that it is computationally inexpensive compared to free energy-based methods. However, we note that the predictions are not consistent for amino acids that are outside the active site, and there are no provisions in the methodology at this moment to address double or triple mutants.