Did you try UCSC genome browser?

Assuming you're looking at human sequences, you can access processed datasets in gtf and bed format at the Broad Institute:

http://www.broadinstitute.org/genome_bio/human_lincrnas/?q=lincRNA_catalog

Feyzollah, I have tried using their genome browser and Ensembl... but no luck. In Ensembl here is an option to blast against ncRNA, but I have never gotten any results back. Even after leaving my computer running for a over 24 hours.

Check out http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE13765 which summarises Guttman et al's paper on lincRNAs (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2754849/)

Hi,

I have been working with lncRNAs for the last year;

Guttman et al papers don't provide transcript structure, but in the 2011 paper they at least provide start/end of studied transcript.

Globally, your best chance is to find Cage data, that would confirm the 5' - since these transcripts are polyAdenylated, 3' is usually a bit more reliable. You can check the chromatin marks peaks used to confirm a 5', if all reads seem to end at the same place, etc. Without chromatin marks at least, I wouldn't trust the 5' end. You can also check Kutter et al 2012 (liver lncRNAs).

If assembled transcripts are not provided, you can use Scripture software (see Guttman 2009)...

If you have assembled transcripts (ie exon structure), you can add them as a track in the UCSC mouse genome browser and therefore check with any of their intersection tools for overlapping existing mRNAs, ESTs, chromatin marks etc. You can also use galaxy (but I would suggest to use join instead of intersect). But anyway, if you have access to linux/unix system, I would advise to use bedtools after downloading sets of interest, it will be faster.

Hope that helps - I never tried to use Encode mouse data!

~A

Thank you Aurelie. I will look into what you've suggested.