Post-translational modification is the attachment of biochemical functional groups after translation from mRNA. Among the different post translational modifications, phosphorylation happens to be one of the most important types which is responsible for important cellular operations. In this research work, we have used multilayer perceptron (MLP) to predict protein residues which are phosphorylated. As features, we have used position-specific scoring matrices (PSSM) generated by PSI-BLAST algorithm for each protein sequence after three runs against 90 % redundancy reduced Uniprot database. For an independent set of 141 proteins, our system was able to provide the best AUC score for 36 proteins, highest for any other predictor. Our system achieved an AUC score of 0.7239 for all the protein sequences combined, which is comparable to the state-of-the art predictors.
CITATION STYLE
Banerjee, S., Ghosh, D., Basu, S., & Nasipuri, M. (2016). JUPred_MLP: Prediction of phosphorylation sites using a consensus of MLP classifiers. In Advances in Intelligent Systems and Computing (Vol. 404, pp. 35–42). Springer Verlag. https://doi.org/10.1007/978-81-322-2695-6_4
Mendeley helps you to discover research relevant for your work.