Bhageerath—Targeting the near impossible: Pushing the frontiers of atomic models for protein tertiary structure prediction#

Abstract

Listen

Translate article

Protein folding, considered to be the holy grail of molecular biology, remains intractable even after six decades since the report of the first crystal structure. Over 70,000 X-ray and NMR structures are now available in protein structural repositories and no physico-chemical solution is in sight. Molecular simulation methodologies have evolved to a stage to provide a computational solution to the tertiary structures of small proteins. Knowledge base driven methodologies are maturing in predicting the tertiary structures of query sequences which share high similarities with sequences of known structures in the databases. The void region thus seems to be medium (>100 amino acid residues) to large proteins with no sequence homologs in the databases and hence which has become a fertile ground for the genesis of hybrid models which exploit local similarities together with ab initio models to arrive at reasonable predictions. We describe here the development of Bhageerath an ab initio model and Bhageerath-H a hybrid model and present a critique on the current status of prediction of protein tertiary structures. Sequence to structure prediction is made using Bhageerath and Bhageerath-H method.