PMC:1892782 / 994-2259
Annnotations
2_test
{"project":"2_test","denotations":[{"id":"17540014-16273071-1690037","span":{"begin":86,"end":87},"obj":"16273071"},{"id":"17540014-16628248-1690037","span":{"begin":86,"end":87},"obj":"16628248"},{"id":"17540014-16204130-1690038","span":{"begin":106,"end":107},"obj":"16204130"},{"id":"17540014-16425273-1690039","span":{"begin":121,"end":122},"obj":"16425273"},{"id":"17540014-17510325-1690040","span":{"begin":779,"end":780},"obj":"17510325"},{"id":"17540014-10329189-1690041","span":{"begin":1219,"end":1220},"obj":"10329189"}],"text":"Introduction\nGenome wide computational screens for structured ncRNA genes in mammals [1-3], urochordates [4], nematodes [5], and drosophilids [6] resulted in tens of thousands putative structured ncRNAs. Functional and structural annotation of these predictions thus becomes a pressing problem. Evidence for evolutionary conservation of RNA structure alone usually does not distinguish very well between the two possible reading directions. This information, however, is crucial already for the most basic annotation information. Direction information is needed e.g. to determine whether a conserved RNA motif is intronic, located within a coding sequence or an untranslated exon, an independent ncRNA, or one of the many classes of small RNAs associated with other transcripts [7].\nThe RNAstrand tool is designed specifically to predict the reading direction of a multiple sequence alignment under the assumption that the alignment contains an evolutionary conserved RNA secondary structure. Our task at hand is a conceptually simple two class prediction problem for which we employ a support vector machine (SVM) [8]. The basic idea is to devise descriptors that utilize both the small asymmetry in the energy rules [9] and the asymmetric effect of GU base pairs."}