| Home | Submit | Dataset | References | Contact |

20S iPred webserver assists in the classification of compounds as 20S proteasome inhibitors or non-inhibitors. Prediction is based on SVM model that has been developed using 64 inhibitors and 208 putative non-inhibitors of 20S proteasome. The non-linear SVM model using RBF kernel showed a cross-validation (CV) accuracy of 97% and Area Under Curve (AUC) of 0.99. Additional validation of the classifier over an independent test dataset showed an accuracy, sensitivity, specificity and Matthews Correlation Coefficient (MCC) of 98%, 93%, 100 % and 0.96 respectively. We have reported the identification of novel drug-like 20S proteasome inhibitors with potential activity against the 20S beta subunit of P. falciparum using a combination of ligand based (Support Vector Machines) and receptor based (molecular docking) techniques. The foussed librabry of compounds developed by the above method is available for download

Related publication:
Subramaniam S, Mehrotra M, Gupta D. Support Vector Machine based prediction of P. falciparum proteasome inhibitors and development of focused library by molecular docking. CCHTS 14(10):898-907(2011)

Bioinformatics Lab, ICGEB, NewDelhi