CnnPOGTP: a novel CNN-based predictor for identifying the optimal growth temperatures of prokaryotes using only genomic<i>k</i>-mers distribution
Shaojing Wang, Guoqiang Li, Zitong Liao, Yunke Cao, Yuan Yun, Zhaoying Su, Xuefeng Tian, Ziyu Gui, Ting Ma
Abstract
SUMMARY: Temperature is very important for the growth of microorganisms. Appropriate temperature conditions can improve the possibility for isolation of currently uncultured microorganisms. The development of metagenomic binning technology had dramatically increased the availability of genomic information of prokaryotes, providing convenience to infer the optimal growth temperature (OGT). Here, we proposed CnnPOGTP, a predictor for OGTs of prokaryotes based on deep learning method using only k-mers distribution derived from genomic sequence. This method was annotation free, and the predicted OGT could be obtained by simply providing the genome sequence to the CnnPOGTP website. AVAILABILITY AND IMPLEMENTATION: http://www.orgene.net/CnnPOGTP. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.