Fig. 9

Overview of the computational pipeline for identifying beaver lncRNAs. Transcript contigs from the consensus transcriptome (“Merged Transcriptome” above) were sequentially filtered using (1) Basic Local Alignment Search Tool for nucleotide sequence (BLASTn) against the NCBI nucleotide database to eliminate probable orthologs of protein-coding genes, known lncRNAs, and other non-lncRNA transcript types; (2) CPAT to detect and eliminate contigs with protein-coding ORFs or nucleotide hexamer usage patterns that are consistent with protein coding genes; (3) HMMscan scan against the Pfam database to identify matches to protein domain motifs; and (4) BLASTn alignment against the OSU draft beaver genome assembly and eliminating those contigs that overlapped with scaffold regions that were annotated (by MAKER) as protein-coding genes. Contigs discovered by the annotation pipeline that are orthologs of known lncRNAs are shown in purple, and novel noncoding contigs identified by the annotation pipeline are shown in green