Change paired end library from FR to RF

Question

[Jake Warner]

Dear all, I have several RNAseq paired end datasets that I want to pool for input into a Trinity de novo assembly. The problem is that half of my datasets are of library type RF (first read (/1) of fragment pair is sequenced as anti-sense (reverse(R)), and second read (/2) is in the sense strand (forward(F))) and half of my data are FR: first read (/1) of fragment pair is sequenced as sense (forward), and second read (/2) is in the antisense strand (reverse).

Is there a way to convert the library type? Simply reverse complementing the reads will not work since then they won't be "pointing" at each other. Any ideas?

Answer 1

[Brian Bushnell]

For the FR pairs, you simply swap R1 and R2, and they become RF. But that's if you know which pairs are FR, which you probably don't. It's not clear why you want to do this, though; Trinity should be able to handle unstranded libraries without a problem.

[h.mon]

Good suggestion. He probably knows which libraries are FR and which are RF, so he can follow your suggestion. His assembly should be a bit better using strand information.

[Brian Bushnell]

Oh, I guess I misinterpreted the question as "half my reads are RF". Yes, if all libraries are purely FR or purely RF, it's easy.

[Jake Warner]

Thanks, Brian. I worked out which libraries were incorrect by aligning to another transcriptome and looking at the flags in the SAM output. With that I could calculate a ratio of "first read mapped to forward strand" to "first read mapped to reverse strand". Your suggestion about swapping the R1 and R2 worked and reversed this ratio so I'm good to assemble!

[Brian Bushnell]

Glad to hear it! If you think this answer resolved the question, please "accept" it so that other people know it is resolved.