SUMOylation Site Prediction using Support Vector Machines

SUMOuylation has been identified as a crucial post-translational modification responsible for many if not all cellular processes. A considerable amount of information pertaining to this process is still elusive. The available information can still be used in development of in-silico procedures to guide the prediction of SUMOuylation sites in protein substrates. Since a highly accurate prediction system is essential to guide efficient experimental designs. We put forth a SUMOuylation site prediction program using support vector machines for discriminating SUMOylating substrates from non-SUMOylating substrates. The algorithm uses manually curated data of experimentally verified SUMOuylation sites as training data. As the number of experimentally verified proteins increases, the accuracy and efficiency of the program will increase accordingly. The web interface for the tool is still under development, meanwhile the stand alone version of the program can be requested from the corresponding author at the given email address.