Diel gene expression improves software prediction of cyanobacterial operons

Cyanobacteria are important participants in global biogeochemical process, but their metabolic processes and genomic functions are incompletely understood. In particular, operon structure, which can provide valuable metabolic and genomic insight, is difficult to determine experimentally, and algorithmic operon predictions probably underestimate actual operon extent. A software method is presented for enhancing current operon predictions by incorporating information from whole-genome time-series expression studies, using a Machine Learning classifier. Results are presented for the marine cyanobacterium . A total of 15 operon enhancements are proposed. The source code is publicly available.