Time Series Analysis of SARS-CoV-2 Genomes and Correlations among Highly Prevalent Mutations
Neha Periwal, Shravan B. Rathod, Sankritya Sarma, Gundeep S. Johar, Avantika Jain, Ravi Pratap Barnwal, Kinshuk Raj Srivastava, Baljeet Kaur, Pooja Arora, Vikas Sood
Abstract
We performed a meta-analysis on SARS-CoV-2 genomes categorized by collection month and identified several significant mutations. Pearson correlation analysis of these significant mutations identified 16 comutations having absolute correlation coefficients of >0.4 and a frequency of >30% in the genomes used in this study. The correlation results were further validated by another statistical tool called hierarchical clustering, where mutations were grouped in clusters on the basis of their similarity. We identified several positive and negative correlations among comutations in SARS-CoV-2 isolates from around the world which might contribute to viral pathogenesis. The negative correlations among some of the mutations in SARS-CoV-2 identified in this study warrant further investigations. Further analysis of mutations such as T85I in nsp2 and Q57H in ORF3a protein revealed that these mutations tend to destabilize the protein relative to the wild type, whereas P323L in RdRp is neutral and has a stabilizing effect. Thus, we have identified several comutations which can be further characterized to gain insights into SARS-CoV-2 evolution.