Predicting Non-protein-coding RNA Genes in Escherichia Coli Using SVM with Signature Descriptor

Non-protein-coding RNA (ncRNA) genes are known to play significant roles. Along with transfer RNAs, ribosomal RNAs and mRNAs, ncRNAs contribute to gene splicing, nucleotide modification, protein transport and regulation of gene expression. Several methods exist for predicting ncRNA genes in Escherichia coli (E.coli). In this paper, we describe a very general, highthroughput method for predicting ncRNA genes in E.coli. The method predicts more than two hundred intergenic regions to contain ncRNA genes, and over half of these overlap with previous tested candidates. Our results indicate that the number of ncRNA genes in E.coli is larger than what has previously been estimated.