Introduction

MHCIPREDS allows users to get predictions from multiple MHC-I predictors, covering the major approaches for predicting MHC-I peptides, in one web server that does not rely on submitting the query sequence to other external MHC-I prediction servers. The server supports predictions using the best performing qualitative and quantitative MHC-I prediction methods considered in our comparison study [EL-Manzalawy and Honavar, 2008]. Specifically, the server supports qualitative MHC-I predictions using modified position specific scoring matrix (MPSSM), AUC optimized matrix method (AOMM), stabilized matrix method for binary classification (SMMBin), and support vector machine (SVM) predictors. For quantitative MHC-I predictions, the server provides predictions using SMM and Gaussian process (GP) predictors. The current implementation of the server provides predictions for the 22 MHC-I HLA alleles. However, we plan to enrich the server with more MHC-I alleles as soon as enough training data become available.

Citation
EL-Manzalawy Y, Honavar V (2009) Qualitative versus quantitative approaches for predicting MHC-I peptides. Submitted.