Skip to content

Latest commit

 

History

History
5 lines (5 loc) · 349 Bytes

README.md

File metadata and controls

5 lines (5 loc) · 349 Bytes

Steps

  • Generative pretraining on SMILE & Amino acids sequences ✅
  • Prove that learnt embeddings are meaningful ✅
  • Determine which models work best on DTI prediction (contrastive leanrning-based, crossentropy based, etc)
  • Interpretability (Attention based models / addition of attention mechanism in alignment of SMILE and amino embeddings)