Summary
Contrastive learning allows functional protein sequence (clusters) to be effectively segregated and improves protein function prediction (1). Segregation of EC numbers in the embedding space was greater compared to the starting (ESM)-1b model.
Figures
Ref (1)
1.
Yu T, Cui H, Li JC, Luo Y, Jiang G, Zhao H. Enzyme function prediction using contrastive learning. Science. 2023;379(6639):1358–63. Available from: https://doi.org/10.1126/science.adf2465