# Enrichment of G‐to‐U Substitution in SARS‐CoV‐2 Functional Regions and Its Compensation via Concurrent Mutations

## TL;DR

This study analyzed SARS-CoV-2 mutations and found that G-to-U substitutions are enriched in functional regions and may be compensated by other mutations, possibly linked to immune responses and disease severity.

## Contribution

The study reveals a compensatory mechanism for G-to-U mutations in SARS-CoV-2 and links them to host immune responses and disease severity.

## Key findings

- G-to-U substitutions are over-represented in functional regions like the S protein and are compensated by concurrent mutations.
- C-to-U mutations are more frequent in unvaccinated individuals and declined until late 2021 before resurging in early 2022.
- Cancer patients showed higher G-to-U frequencies, suggesting a potential link to oxidative stress and reactive oxygen species.

## Abstract

We surveyed single nucleotide variant (SNV) patterns from 5 903 647 complete SARS‐CoV‐2 genomes. Among 10 012 SNVs, APOBEC‐mediated C‐to‐U (C > U) deamination was the most prevalent, followed by G > U and other RNA editing‐related substitutions including (A > G, U > C, G > A). However, C > U mutations were less frequent in functional regions, for example, S protein, intrinsic disordered regions, and nonsynonymous mutations, where G > U were over‐represented. Notably, G‐loss substitutions rarely appeared together. Instead, G‐gain mutations tended to more frequently co‐occur with others, with a marked preference in the S protein, suggesting a compensatory mechanism for G loss in G > U mutations. The temporal patterns revealed C > U frequency declined until late 2021 then resurged in early 2022. Conversely, G > U steadily decreased, with a pronounced drop in January 2022, coinciding with reduced COVID‐19 severity. Vaccinated individuals exhibited a slightly but significantly higher C > U frequency and a notably lower G > U frequency compared to the unvaccinated group. Additionally, cancer patients had higher G > U frequency than general patients during the same period. Interestingly, none of the C > U SNVs were uniquely identified in 2724 environmental samples. These findings suggest novel functional roles of G > U in COVID‐19 symptoms, potentially linked to oxidative stress and reactive oxygen species, while C > U remains the dominant substitution, likely driven by host immune‐mediated RNA editing.