Abstract
The massive amount of data generated from genome sequencing have given rise to several mutation predictor tools although no mutation database or predictor tool have been developed specifically for the transmembrane region of membrane proteins.
We present TMSNP, a database that currently contains information from 2624 pathogenic and 195964 non-pathogenic reported mutations located on the TM region of membrane proteins. The computed conservation parameters and annotations on these mutations were used to train a machine-learning model that classifies TM mutations as pathogenic or non-pathogenic. The presented tool improves considerably the prediction power of commonly used mutation predictors and additionally represents the first mutation prediction tool specific for TM mutations.
TMSNP is available at http://lmc.uab.es/tmsnp/
Contact mireia.olivella{at}esci.upf.edu