A greater proportion of viral sequences are closely related to one another than prokaryotic and eukaryotic sequences, and these sequences get removed during clustering

Summary

A greater proportion of viral sequence closely co-cluster together than prokaryotic and eukaryotic sequences, and these sequences get lost during (clustering) (1). For example, 90% clustering reduces prokaryotic/eukaryotic sequences 8-fold but viral sequences 550-fold.

Figures

Ref (1)

1.

Gurev S, Youssef N, Jain N, Mehrotra A, Leung SRM, Jackson A, et al. Evaluating variant effect prediction across viruses. openRxiv; 2025. Available from: https://doi.org/10.1101/2025.08.04.668549