[cjp]

Hi adameur,

I'd thought to mention your paper in my previous answer as well!

Do you know how much it may affect mRNA-seq studies - where people try to select processed RNA's from total RNA using poly-A and related lab methods?

Chris

[adameur]

Hi Chris,

From what I have seen the choice of lab method makes a big difference.. PolyA+ gives a much higher proportion of reads mapping to exons, which is good if you're mainly interested in expression of mature mRNAs and splice variants.

Total RNA on the other hand makes it possible to also study other types of events, like nascent transcript formation and co-transcriptional splicing. The drawbacks are that you need quite a lot of reads to study the intronic expression and that there are no established analysis tools.

Adam

[polsum]

a dumb question. How did you make that figure in the first post of this thread? which program you used?

[Simon Anders]

Just to clarify: We have two things going on here.

(i) The coverage curve shown in post #1, produced with samtools, disagrees from the one shown in post #3, produced, for the same data, with IGC. This points to an interesting and worrying bug in samtools and explains most of the coverage in the introns.

(ii) There are still reads left in the introns. They are most likely really there, and many people have seen them in RNA-Seq data. The thread has probably mentioned most of the suggestions the literature has discussed recently about what this means.

[cjp]

Yes is a two-issue thread!

The samtools thing is that mpileup adds '>' and '<' symbols where the cigar part of an alignment is an N. You can use "samtools depth" and that gives me 0 over introns.