UbPred is a random forest-based predictor of potential ubiquitination sites in proteins. It was trained on a combined set of 266 non-redundant experimentally verified ubiquitination sites available from our experiments and from two large-scale proteomics studies (Hitchcock, et al., 2003; Peng, et al., 2003). Class-balanced accuracy of UbPred reached 72%, whereas the AUC (area under the ROC curve) was estimated to be ~80%.

Supplementary information, usage instractions and more about the datasets can be found on the help page.

UbPred executable files are available for download:

Installation guide can be downloaded here.

  Pre-computed ubiquitination predictions on the whole yeast proteome can be downloaded here.

UbPred was developed by Predrag Radivojac (Northeastern University, Khoury College of Computer Sciences), Vladimir Vacic (Columbia University) and Lilia Iakoucheva (University of California, San Diego).

In citing the UbPred software, please refer to:

Radivojac, P., Vacic, V., Haynes, C., Cocklin, R. R., Mohan, A., Heyen, J. W., Goebl, M. G., and Iakoucheva, L. M. Identification, Analysis and Prediction of Protein Ubiquitination Sites. Proteins: Structure, Function, and Bioinformatics. 78(2):365-380. (2010)

Please direct all comments and suggestions to predrag@northeastern.edu or lilyak@ucsd.edu.