ncRNA discovery and functional identification via sequence motifs

Non-coding RNAs play regulatory roles in gene expression via establishing stable joint structures with target mRNAs through complementary sequence motifs. Sequence motifs are also important determinants of the structure of ncRNAs. Here we introduce two computational tools that both exploit differential distributions of short sequence motifs in ncRNAs for the purpose of identifying their loci and functionality. (1) smyRNA is an ncRNA search tool which is based on identification of "structural" motifs on a genome sequence. (2) pRuNA is the basis of an index structure which can search a collection of ncRNAs to determine those that can establish stable interactions with a query mRNA. pRuNA uses differential distribution of "functional" motifs for the purpose of eliminating a significant majority of the ncRNAs while retaining the actual regulatory RNAs of a query mRNA. The retained ncRNAs are then verified by inteRNA, a fast technique for determining the topology of the joint secondary structure between an ncRNA and its potential target. In cross-training experiments on the e.coli genome, smyRNA identified the majority of the test ncRNAs while yielding no false positives. It also identified a number of previously unknown ncRNA candidates which are likely to be functional. We also tested our index structure on the e.coli ncRNAs: for a typical mRNA query, pRuNA eliminated 91% of the index while retaining the known regulatory RNA. On the complete Rfam (seed) ncRNA collection, the pruning efficiency of pRuNA was 86%. In both cases, the interaction between each query mRNA and its known regulatory RNA was verified by inteRNA.