The abstract is read at the publisher; the summary is JClub's.
antiGen forecasts SARS-CoV-2 spike protein mutations from phylogenetic trees with state-of-the-art accuracy, including receptor binding domain changes, unseen mutations, and high-prevalence variants.
- Why it matters: Predicting future viral genotypic change from historical phylogenies is difficult, yet it is crucial for pandemic preparedness and for designing interventions and therapeutics.
- What they did: antiGen is a machine learning model trained on evolutionary histories to forecast future amino-acid changes, evaluated on SARS-CoV-2 spike and also influenza, respiratory syncytial virus, and dengue surface proteins.
- The result: Experimentally tested antiGen-forecasted spike variants largely retain pseudoviral infectivity in vitro, and a model trained across 49 viral proteins from 27 viruses improves performance versus single-protein phylogeny models.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.