Multimodal Representation Learning for Proteins

  • Developing multi-modal (ie. sequence, structure) representation learning pipeline for protein fitness (ie. binding, catalysis) prediction
  • Building easy-to-use protein single- to multi-mutant fitness prediction by incorporating mutation sequence context (coevolution, stability, and biochemical rules) on 11 diverse datasets
Francesca-Zhoufan Li
Francesca-Zhoufan Li
AI for Science & Engineering, currently focusing on machine learning for proteins

I apply machine learning to science and engineering problems with a current focus on protein engineering.