Reasoning for detection and benchmarking of viral protein structural mimicry

This reasoning chain explains how the pipeline combines experimentally validated viral mimics, predicted structures from Viro3D/AlphaFoldDB, Foldseek structural comparisons, and Bayesian GMM clustering to reliably detect both known mimics and inform threshold selection in ambiguous cases. It connects the benchmarking evidence to the overarching methodological claim that the pipeline enables systematic discovery of protein structural mimicry.

Reasoning Steps (2)

Source

Connections (1)