Pre-trained molecular representations enable antimicrobial discovery
Pre-trainingDataset: The MolE pre-training dataset was created by randomly sampling 100,000 unlabeled molecules from a collection of 10,000,000 unique structures...
antimicrobial
Pre-trainingDataset: The MolE pre-training dataset was created by randomly sampling 100,000 unlabeled molecules from a collection of 10,000,000 unique structures...
