• Expand collection by over 45%, improved functional biological activity tree (managed by managed by PostgreSQL)
• Train over 90 binary classification models using protein language models and machine learning. Over 100,000 peptides with biological activities and 3.8 million sequences with predicted activities