Learning the language of proteins

Presenter Details

Dr. Claire McWhite, assistant professor, molecular and cellular biology

I’m interested in the acquisition of novel protein function across evolution. It is possible to model protein sequences as a language, viewing amino acids as words and whole proteins as sentences. As in human language, where words have meaning based on context, amino acids have function based on the context of their surrounding amino acids. These models are a powerful tool to interpret both mutation effects and protein functional divergence. I’m broadly focused on developing frameworks to interpret biological properties through the lens of protein language.

Areas of research: Systems biology, protein language models, evolution, plants, proteomics

Presentation Details

We study the fundamental connection between protein amino acid sequences and protein function. By conceptualizing protein sequences as a language, where each amino acid represents a unique "word," we study the intricate "grammar" that governs how these sequences encode physical properties and functional information. Our work introduces novel techniques for modifying protein properties, examining the sequence determinants of functional variation within highly expanded protein families, and understanding the specific functional effects of sequence variation.