Predicting Transmissibility-Increasing Coronavirus (SARS-CoV-2) Mutations

Abstract

Advantageous variants of the SARS-CoV-2 virus have arisen through mutations, particularly on a single amino acid basis. These point mutations can cause changes in the structure of SARS-CoV-2 and affect the efficiency of interaction with the ACE2 protein. N501Y and E484K mutations affecting binding by ACE2 have been widely observed. This study aimed to predict SARS-CoV-2 mutations that could be as effective as N501Y and E484K and pose a danger due to their high contagiousness. Experimental data on SARS-CoV-2 and ACE2 binding and stability were associated with different amino acid properties and integrated into machine learning and computational biology techniques. As a result of the analyses made in algorithms, N501M, Q414A, N354K, Q498H and N460K have been predicted to be likely to have a dangerous effect. The N501W mutations are most likely to have dangerous effects on the spread of the coronavirus. We suggest that attention should be paid to the position 501 mutation since this position is repeated in the lists of mutations that the algorithm detected as dangerous. G446, G447, Y505, T500, Q493, Y473, and G476 were determined as the positions where dangerous variants could be seen as a result of the analyses of the multiple interaction data created with the ACE2 and RBD interaction data. The 13 dangerous positions and mutations have been detected to accurately describe the position of the mutations caused by the Omicron variant and were among the known dangerous mutations similar to those occurring at Q498, G446, Y505 and Q493 positions.

Keywords

• coronavirus
• COVID-19 pandemic
• prediction algorithm
• mutation
• bioinformatics