The DELPHI Web Server
Department of Computer Science
University of Western Ontario
Contact information:
Yiwei Li: yli922uwo.ca
Lucian Ilie: ilieuwo.ca

The Delphi server is no longer maintained, but you can still access the codes at GitHub. We encourage you to explore and utilize our more recent algorithm, which can be found at Seq-InSite.

Downloads

Training and testing datasets used in DELPHI can be downloaded here.
The prediction results of DELPHI and other programs on the Dset_448, Dset_355, Dset_72, Dset_186, and Dset_164 are available here.
The datasets and DELPHI predictions for sites conservation are available here.
The human proteins in fasta format are available here.
The DELPHI predicted human proteome is available here.

Source code

The DELPHI web server allows up to 10 sequences per submission. It is recommended to use the source code for high-throughput predictions. The DELPHI source code is freely available on github.

Citation

- Li, Yiwei, Golding, Brian, and Ilie, Lucian, DELPHI: accurate deep ensemble model for protein interaction sites prediction. Bioinformatics, 2020, btaa750
- Li, Y. and Golding, G. Brian and Ilie, L., DELPHI: accurate deep ensemble model forprotein interaction sites prediction (ISMB’20), 3DSIG, Montreal, 2020. (abstract and poster )