Provide a knowledge-based fragment binding predictor as well as the underlying knowledge base.
This project is a data-driven approach to predicting small molecule fragments that will bind a protein structure of interest. There are two parts to the project: (1) the knowledge base contains all known protein-fragment interactions from the Protein Data Bank (2) the method uses the knowledge base to make fragment predictions given an input structure. Each fragment prediction yielded by the method is accompanied by all supporting pieces of information from the knowledge base to facilitate user interpretation of the results.
The knowledge base of protein-fragment interactions will be released following publication.
The code base (in python) for predicting fragments using the knowledge base will be released following publication.