The program PhyloCSF (9/16/2010 release) [40] was used to computationally evaluate the coding potential of the filtered lincRNA transcripts. A BED file describing these lincRNA transcripts as well as a random sample of 8310 RefSeq NM transcripts was loaded onto the Galaxy webserver (https://main.g2.bx.psu.edu/) and the tool ‘Stitch Gene Blocks’ was used to retrieve multiple alignment files with sequence entries for the following genome builds based on the 44 way Multiz alignment to hg18: hg18 panTro2 rheMac2 tarSyr1 micMur1 otoGar1 tupBel1 mm9 rn4 dipOrd1 cavPor3 speTri1 oryCun1 ochPri2 vicPac1 turTru1 bosTau4 equCab2 felCat3 canFam2 myoLuc1 pteVam1 eriEur1 sorAra1 loxAfr2 proCap1 echTel1 dasNov2 choHof1. Genome build names were converted to common names and PhyloCSF was run using the options –orf = StopStop3 and –frames = 6.
