Bhumika Singh

The paper introduces an active learning strategy using Thompson sampling to efficiently screen peptide libraries with AlphaFold-based binding predictions, achieving significant improvements over random sampling.

This study benchmarks eleven force fields on twelve peptides to assess their ability to model folding behaviors, revealing structural biases and limitations in capturing disorder and secondary structure.

MELD-Adapt dynamically evaluates data reliability to improve structure prediction accuracy by adjusting initial beliefs and identifying reliable data fractions.

This paper introduces a computational pipeline using AlphaFold2 to prioritize protein-protein binding candidates from high-throughput libraries and identify binding epitopes.