N. J. Rollins

Protein design using structure-based residue preferences

This paper introduces a method to design functional proteins using residue preferences derived from structural data, showing high accuracy in predicting combinatorial mutation effects.

Learning from pre-pandemic data to forecast viral escape

This paper develops EVEscape, a framework predicting viral escape mutations using historical data and structural information, showing it can forecast SARS-CoV-2 variants as effectively as experimental methods.