Genome-wide identification and prediction of SARS-CoV-2 mutations show an abundance of variants: Integrated study of bioinformatics and deep neural learning

Abstract

Genomic data analysis is a fundamental system for monitoring pathogen evolution and the outbreak of infectious diseases. Based on bioinformatics and deep learning, this study was designed to identify the genomic variability of SARS-CoV-2 worldwide and predict the impending mutation rate. Analysis of 259044 SARS-CoV-2 isolates identify 3334545 mutations (14.01 mutations per isolate), suggesting a high mutation rate. Strains from India showed the highest no. of mutations (48) followed by Scotland, USA, Netherlands, Norway, and France having up to 36 mutations. Besides the most prominently occurring mutations (D416G, F106F, P314L, and UTR:C241T), we identify L93L, A222V, A199A, V30L, and A220V mutations which are in the top 10 most frequent mutations. Multi-nucleotide mutations GGG>AAC, CC>TT, TG>CA, and AT>TA have come up in our analysis which are in the top 20 mutational cohort. Future mutation rate analysis predicts a 17%, 7%, and 3% increment of C>T, A>G, and A>T, respectively in the future. Conversely, 7%, 7%, and 6% decrement is estimated for T>C, G>A, and G>T mutations, respectively. T>G\A, C>G\A, and A>T\C are not anticipated in the future. Since SARS-CoV-2 is evolving continuously, our findings will facilitate the tracking of mutations and help to map the progression of the COVID-19 intensity worldwide.