Welcome to GeoNet!

        Predicting interaction sites between proteins and other biomolecules, such as nucleic acids and other proteins, is essential for most biological processes. GeoNet is a powerful and interpretable geometric deep learning model for identifying protein interaction sites based on protein structure.

        GeoNet constructs a local sphere containing geometric information that is centered on the target amino acid and includes other residues within a fixed radius for each residue in a protein.

        Users can GeoNet to predict the DNA, RNA, and protein binding sites of a protein through our online GeoNet webserver. You can also use the source code of GeoNet which is provided on our Github page http:.

If you think GeoNet is useful, please kindly cite the following paper:

Webserver update: