Xuenan Mi

Variational Autoencoder-enabled High-throughput Drug Screening for HIV Latency Modulators predicted through Noise in Gene Expression

This study uses a variational autoencoder to screen ~175,000 compounds for HIV latency modulators, identifying 16 LRAs and 2 LPAs with a 15.9% experimental hit rate.

De novo Folding Mechanisms of Lasso Peptides

This study uses molecular dynamics and deep learning to uncover the folding mechanisms of lasso peptides, revealing key thermodynamic and kinetic principles.

LassoESM: A tailored language model for enhanced lasso peptide property prediction

This study develops LassoESM, a language model tailored for lasso peptide property prediction, showing improved performance in substrate compatibility and biological activity prediction.

Deep mutational scanning reveals sequence to function constraints for SWEET family transporters

This study uses deep mutational scanning to analyze sequence-function relationships in the plant SWEET transporter AtSWEET13, identifying mutations that enhance transport function and revealing constraints in transmembrane helix 4.