PDZ domain-peptide interaction prediction website


POW is a website that allows users to predict PDZ domain-peptide interactions for human, mouse, worm and fly PDZ domains. Predictions are made using a support vector machine (SVM) that was trained using experimentally determined PDZ interaction data from protein microarray and phage display experiments for mouse and human [1,2]. Two types of predictors are available for use. The first is sequence-based (trained using domain and peptide sequence features) while the other structure-based (trained using domain structure and peptide sequence features).

Predict Interactions

To predict interactions, select an organism (default: Human) and select a predictor (default: Sequence-based). Next select PDZ domain(s) from the drop down list and input your peptides (i.e. upload a file or type them in the textfield). Click the Submit button.

Predicted hits will be displayed in a new window with the domain name, binding peptide sequence, SVM decision score and Ensembl protein ID. The SVM decision score is a measure of predictor confidence in the prediction. Ensembl protein ID will only be shown if the predicted binder is genomic (according to genome assembly GRCh37.64 for human, NCBIM37.64 for mouse, BDGP5.25.64 for fly and WS220.64 for worm).


Shirley Hui (shirley.hui *a-t* utoronto.ca)

Xiang Xing (xiang.xing *a-t* utoronto.ca)

Gary D. Bader (gary.bader *a-t* utoronto.ca)


  1. Proteome scanning to predict PDZ domain interactions using support vector machines, Hui S, Bader GD, BMC Bioinformatics 2010, 11:507 (12 October 2010) Paper, Supplementary Information
  2. Predicting PDZ domain mediated protein interactions from structure, Hui S, Xing X, Bader GD, (Feb 2012 Submitted) Paper, Supplementary Information