Prediction of protein structure from invariant fragment configurations

Abstract

Proteins are ubiquitous in all living organisms be it unicellular or multicellular. Usually In order to determine the function , protein structure has to be known. Existing practice of protein structure elucidation/determination relies largely upon sophisticated, complex, expensive, laborious, time consuming and highly skill demanding instrumental methods like NMR, X-Ray Crystallography, cryo-electron microscopy. The problem is compounded when it becomes impossible to have proteins in crystallized form for in many cases this crystallization itself is impossibility. It is for the above reasons that computational methods are preferred since impediments could be circumvented. Hence, protein structure prediction by computational methods has become attractive. There are many established methods for protein structure prediction like Comparative modeling and ab initio methods but these methods fall short of predicting all the protein structures that can exist in nature. Further, all these methods use some sort of energy minimization techniques to arrive at the structure. We know that protein structure is governed by the amino acid sequence. Therefore considering amino acid sequences as the soul, determining the sequences which have invariant configurations of protein fragments and compiling them in a database and predicting the protein structures of determined and yet to be determined is ultimate goal of this research undertaken. The attempt made in this work is to determine the fragments of existing and known sequences of amino acids whose structure is already known which are unique in themselves and contribute in making up the structure. Developing, analyzing and compiling such protein sequence fragments of particular(unique) length which are invariant both in sequence identity and configuration identity is a key step in developing a database. The database so developed will later have the fragments whose positions of atoms in the three dimensional space does not change or vary significantly.