recommendations on paired-end analysis?

[Zhuzhu]

Hi everyone,

I am pretty new to the RNA-seq, and currently looking at the paired-end data. Our primary goal is to find fusion proteins. Any suggestion on which package to use? Or any relevant papers?

Thanks in advance!

[apfejes]

I don't know if there's a single given software package for what you're doing. There are several people at my institution who are working on different methods of interpreting paired end data for fusion proteins.

Unfortunately, It's not a solved problem, as far as I know. (If there's a paper claiming to have solved it completely for all read lengths, I'd love to hear about it!)

You may have to write your own software, or find someone who's already working on this and collaborate with them.

Anthony

[Zhuzhu]

Hey Anthony,

Thanks!

I guess I'd have to work out my own software. If it's possible, would you please share some of the names you know who are currently working on this issue? Thanks a lot again!

Zhuzhu

[swbarnes2]

I think you might have to do it more or less manually. Align each end separately, and then look through the list for clusters where the two ends fall in wildly different places.

Bowtie reports sub-optimal hits, this might help reduce the false positive rate.

[Zhuzhu]

Thanks, swbarnes2!

Yeah, I am now pulling out the paired end reads that are mapped to two different places. And as you indicate, the false positive is a problem.

I haven't tried Bowtie, though planed to do so. It seems I really should. Thank you for the suggestion!