A Peptides Prediction Methodology with Fragments and CNN for Tertiary Structure Based on GRSA2

Proteins are macromolecules essential for living organisms. However, to perform their function, proteins need achieve Native Structure (NS). The NS is reached fast in nature. By contrast, silico, it obtained by solving the Protein Folding problem (PFP) which currently has a long execution time. PFP computationally an NP-hard and considered one of biggest current challenges. There several methods following different strategies PFP. most successful combine computational biological information: I-TASSER, Rosetta (Robetta server), AlphaFold2 (CASP14 Champion), QUARK, PEP-FOLD3, TopModel, GRSA2-SSP. first three named highest quality at CASP events, all apply Simulated Annealing or Monte Carlo method, Neural Network, fragments assembly methodologies. In present work, we propose GRSA2-FCNN methodology, assembles applied peptides based on GRSA2 Convolutional Networks (CNN). We compare with best state-of-the-art algorithms PFP, such as Rosetta, AlphaFold2, Our methodology dataset 60 achieves performance tested common metrics TM-score, RMSD, GDT-TS area.