Multimodal Deep Learning for Protein Engineering

Explore the cutting-edge applications of multimodal deep learning in protein engineering through this comprehensive lecture. Delve into the use of sequences, structures, and biophysical features for predicting protein function and generating functional proteins. Learn about innovative approaches such as machine translation for signal peptide generation, FoldingDiff for protein structure prediction, discrete diffusion mutations, and OA-ARDM for expanding protein function space. Gain insights into the potential of machine learning to revolutionize protein engineering for therapeutic and industrial applications. Engage with Q&A sessions that address key questions in the field.

- intro
- Using Multiple Data Modalities Discover and Design Proteins
- Signal Peptides Simplify Protein Production
- Machine Translation for SP Generation
- Q+A
- Expanding the Function Space
- FoldingDiff
- Discrete Diffusion Mutations
- OA-ARDM
- Can we Generate Proteins With New Functions?
- Q+A