Chunk #41 — Materials and Methods — LincRNA Discovery — Transcripts were filtered to remove overlap with non-lincRNA genes or pseudogenes and short transcripts
Transcripts less than 200 nt in length were removed. Remaining transcripts were removed if they were within 1 kb of RefSeq NM genes on the same strand or, in the case of transcripts with ambiguous strandedness, on either strand relative to the NM gene. Transcripts on the opposite strand of an NM gene were removed if they overlapped the NM gene by at least one base. In addition, transcripts overlapping at least one base of any of the following were removed, regardless of strandedness: Ensembl v61 genes except “lincRNA” and “processed_transcript”, non-human RefSeq genes aligned to hg18 with BLAT (UCSC Genome Browser “Other RefSeq” track), alternative and extended 5′ and 3′ UTRs of known human genes from UTRdb, RefSeq NR and XR transcripts annotated as “pseudogenes”, and Ensembl v54 coding sequences.
