PROTEIN STRUCTURE PREDICTION

The present subject matter discloses prediction of the structure of a protein based on an invariant data. For this, an Invariant Fragment library having fragments of length varying from 3 to 41 amino acids from a number of proteins with known 3-D structural data values is created, where each element of the invariant fragment library corresponds to a unique amino acid sequence. The optimized structure invariant data base is held in a suitable storage in a computing system environment supported by one or more processors. Predicting the 3-D structure is carried using one or more processor assisted processing steps that cleave the unknown protein to fragments at multiple tracked points and select 3-D structural data values from the loaded Invariant fragment library and concatenating the final structure removing the overlap of the cleaved fragments.