Accurate predictions of SARS-CoV-2 infectivity from comprehensive analysis

Area of Science:

• Virology
• Molecular Biology
• Computational Biology

Background:

• Accumulated SARS-CoV-2 data offers unprecedented insights into viral evolution.
• Understanding SARS-CoV-2 evolution is crucial for evaluating its infectivity.

Purpose of the Study:

• To investigate SARS-CoV-2 evolutionary features and their impact on infectivity.
• To identify specific amino acid substitutions associated with increased viral infectivity.

Main Methods:

• Analysis of viral sequences to identify amino acid substitutions in the receptor binding motif (RBM).
• Development of the APESS model to evaluate infectivity based on biochemical and mutational properties.
• In silico and in vitro validation of the APESS model.
• Machine learning for predicting potentially prominent mutations.

Main Results:

• Increased frequency of substitutions to lysine (K) and arginine (R) in Variants of Concern (VOCs).
• Common mutations became fixed in the Omicron variant.
• Specific mutations impacted SARS-CoV-2 binding affinity to the ACE2 receptor.
• APESS model demonstrated accuracy in predicting infectivity.

Conclusions:

• Established a link between specific viral properties and increased SARS-CoV-2 infectivity.
• Developed predictive models and tools (APESS, AIVE) for assessing viral infectivity.
• Enhanced understanding of SARS-CoV-2 evolution and improved prediction of viral behavior.