CnMAPKPred: A machine learning approach for predicting mitogen-activated protein kinase (MAPK) in coconut

Abstract

Mitogen-activated protein kinases (MAPK) comprise a large group of ubiquitous proline-directed, protein-serine/threonine kinases. MAPK lies in protein kinase cascades and MAPK pathways are highly conserved signaling pathways that regulate diverse cellular functions including cell proliferation, differentiation, migration and apoptosis. The MAPKs sequenced so far share approximately 40% amino acid sequences. In this study, an attempt has been made in developing a tool for the prediction of coconut MAPKS using machine learning approaches. Of the different algorithms tested, the Naïve Bayes algorithm gave the best results.  To evaluate the prediction performance of the developed algorithm, cross validation and independent data set validation were carried out. The results revealed that the proposed algorithm could be very effective in the computational prediction of MAPK in coconut. This tool is freely available at http://210.212.229.52/cnmapk.