Welcome to SpatPPI!

Protein-protein interactions (PPIs) that occur within intrinsically disordered regions (IDRs) play crucial roles in various biological processes. SpatPPI is a powerful and interpretable geometric deep learning model for identifying intrinsically disordered protein interaction based on protein structure.

SpatPPI predicts protein structure based on AlphaFold2 prediction for prediction, which enhances the understanding of protein spatial conformation by dynamically adjusting the initial input structure and capturing residue contact information.

Users can use the online service to predict intrinsically disordered protein-protein interactions or general protein interactions; SpatPPI primarily predicts human protein-protein interactions but also supports cross-species predictions. SpatPPI webserver. You can also use the source code of SpatPPI which is provided on our Github page https://github.com/Zamprini/SpatPPI.

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

SpatPPI: a geometric deep learning model for predicting protein-protein interactions involving intrinsically disordered regions

Webserver update:

Oct, 9th, 2024: the first version of SpatPPI server was established.

This work is openly licensed via CC0 1.0