Beth-2 Framework Predicts Key SARS-CoV-2 Mutations Nine Months Ahead
Researchers created a computational tool analyzing intra-host sequencing to forecast dominant viral mutations before widespread transmission.
Researchers developed a computational framework named Beth-2 to forecast key evolutionary mutations of SARS-CoV-2, according to a study published in the Proceedings of the National Academy of Sciences. The method leverages within-host viral genome deep sequencing data to detect early fitness signals of variants before they establish widespread circulation at the population level pnas.org
Proceedings of the National Academy of Sciencespnas.orgOpen source ↗.
Retrospective Accuracy and Lead Time
During retrospective validation on genomic datasets, Beth-2 successfully identified 98.6% of key SARS-CoV-2 mutations located on the viral receptor-binding domain. The system detected these evolutionary shifts with a median lead time of 9 months prior to the variants reaching predominance among host populations. The algorithm correlates intra-host evolutionary dynamics with between-host fitness patterns to assess emerging changes.
Prospective Performance and Lineages
In prospective testing involving 26 predicted viral mutations, 69% demonstrated significantly enhanced transmissibility. Five of these mutations exhibited notable immune evasion against blood plasma drawn from recipients of COVID-19 mRNA vaccines six months after the forecast date.
Among the functionally enhanced variants identified by the model was F456L, a defining mutation present in multiple lineages, including KP.3, NB.1.8.1, and XFG. The authors reported that this computational approach captures pathogen mutations under host immune pressure and can be integrated into existing genomic surveillance networks to inform vaccine design pnas.org
Proceedings of the National Academy of Sciencespnas.orgOpen source ↗.
