Physicochemical properties, indices and descriptors for amino-acid sequences.
Active learning workflow to train and fine-tune molecular property predictors with chemist feedback for goal-oriented molecule generation.
Estimate maximum performance bounds based on experimental errors for ML datasets
