Evolutionary annotation of conserved long non-coding RNAs in major mammalian species

Mammalian genomes contain tens of thousands of long non-coding RNAs(lnc RNAs) that have been implicated in diverse biological processes. However, the lnc RNA transcriptomes of most mammalian species have not been established, limiting the evolutionary annotation of these novel transcripts. Based on RNA sequencing data from six tissues of nine species, we built comprehensive lnc RNA catalogs(4,142–42,558 lnc RNAs) covering the major mammalian species. Compared to protein-coding RNAs, expression of lnc RNAs exhibits striking lineage specificity. Notably, although 30%–99% human lnc RNAs are conserved across different species on DNA locus level, only 20%–27% of these conserved lnc RNA loci are detected to transcription, which represents a stark contrast to the proportion of conserved protein-coding genes(48%–80%). This finding provides a valuable resource for experimental scientists to study the mechanisms of lnc RNAs. Moreover, we constructed lnc RNA expression phylogenetic trees across nine mammals and demonstrated that lnc RNA expression profiles can reliably determine phylogenic placement in a manner similar to their coding counterparts. Our data also reveal that the evolutionary rate of lnc RNA expression varies among tissues and is significantly higher than those for protein-coding genes. To streamline the processes of browsing lnc RNAs and detecting their evolutionary statuses, we integrate all the data produced in this study into a database named Phylo NONCODE(http://www.bioinfo.org/phylo Noncode). Our work starts to place mammalian lnc RNAs in an evolutionary context and represent a rich resource for comparative and functional analyses of this critical layer of genome.