Understanding evolution in viruses is vital to identify factors that drive their emergence, adaptability, and pathogenicity. This project focuses on studying tempo (substitution rate with clock behaviour) and mode (selection pressure) of viral evolution across a wide range of viral taxonomy, particularly at the family and genus levels. Viral sequences were downloaded from GenBank and subjected to an automated parsing process. Datasets were then passed through stringent filtering criteria to ensure the production of high-quality alignments suitable for evolutionary analysis.
Temporal signal was measured using molecular clock methods to assess evolutionary rate estimates. 59% of alignments analysed showed sufficient temporal signal, allowing them to progress to substitution rate analysis. Substitution rate was measured using Bayesian frameworks with the best clock model (relaxed vs. strict clock) according to branch variation. The majority of alignments fell into moderate and fast evolving categories, while very slow and very fast were the lowest. According to temporal signal and substitution rate analysis findings, tempo did not align with viral taxonomy levels.
To study evolution mode, selection pressure was analysed over viral alignments using likelihood-based codon models, with 60% of alignments showing positive selection. Functional domains and GO annotations linked the majority of positive selected sites associated proteins to structural and replication processes. Contrary to common belief that selection in viruses is mainly driven by immune evasion, high number of positive selected sites were identified in non-surface proteins. Structural modelling further demonstrated how positive selection can impact protein surfaces and molecular interactions.
Results can contribute to an understanding how viruses evolve in general. By comparing tempo and mode across taxonomic levels, the project aims to evaluate evolutionary patter consistency within different taxonomies. Across all analyses, findings indicated that evolutionary dynamics do not uniformly follow taxonomic classification.