Subu K. Subramanian

This study designs and validates protease-resistant antiviral peptides that inhibit SARS-CoV-2 fusion, showing improved efficacy in vitro and in vivo.

The paper identifies and redesigns energetically frustrated regions in a defective T4 clamp loader to restore function through conformational balance and stability.

The paper fine-tunes language models on laboratory data to design antibodies with improved affinity, achieving sub-25 picomolar binding for four antibodies.

This study uses deep generative models to design synthetic SH3 domains functional in yeast stress signaling, showing a VAE can recapitulate natural function and classify orthologs by phylogeny and function.

This paper shows that deep learning language models can generate functional artificial proteins across families, with activities similar to natural proteins and structures resembling natural folds.