hi, guys, I am working on riboseq experiment. I just got my pair-end sequencing result. And each sample was got four r1 and four r2 reads. My PI told me that I only need working on r1 reads for triming and mapping because R1 reads has all the rna fragments informatiom. Are your guys also only look at r1 reads? I have a script for triming R1 reads. May I use same script for trim R2 reads? I thought I can increase the reads depth by mapping reads to genome using bowtie pair-end alignment. But maybe I am wrong. could you give me some suggestions?
My PI told me that I only need working on r1 reads for triming and mapping
Ask your PI why until you understand why this is being asked of you. Check with them why they went for PE sequencing if they don't even want the R2. It's your job to understand the problem you're dealing with, and the best person to talk to is your PI.
Is your PI a bioinformatician? I have witnessed wet-lab PIs with zero data analysis experience trying to steer bioinformatics projects! I would talk to the facility and understand the rationale behind your PI's comment. @lieven.sterck gave an example of possible a scenario in their answer.
no, if you do PE sequening you analyse both reads, and moreover you analyse them in pair! (don't to trimming on R1 and then on R2 and then combine again .... recipe for disaster ;) )
that being said: there is technical exception on the above, and this is when your sequencing was done on an illumina NextSeq (if I recall correctly). This is because there seem to be issue or unavailability of reagents kit for the SE sequencing mode and as such many seq-providers use the PE kit and only run for a few bases on R2 for technical reasons. In that case, and only in that case (and given that the provider has actually provided them 'cus they usually don't), you can ignore the R2 read (it will anyway only be like 7 bases or such). If both reads have reasonable lengths it's truly PE data and you should process it as such.
technically you could only analyse the R1 read but there is not reason why you would do so (it only has advantages analysing them as PE). The reason you PI gives makes no sense in my opinion.
You will not increase read depth by aligning them as PE , what you will gain is specificity (== your reads should align more accurately/uniquely)
And do in any case also take RamRS advice in account.
Ask your PI why until you understand why this is being asked of you. Check with them why they went for PE sequencing if they don't even want the R2. It's your job to understand the problem you're dealing with, and the best person to talk to is your PI.
My PI have no idea about that. Our facility usually do PE sequencing for us. There are so different kinds of samples sequenced in one run.
could make sense they do so, especially if they pool samples in a run, but also see my answer below
Is your PI a bioinformatician? I have witnessed wet-lab PIs with zero data analysis experience trying to steer bioinformatics projects! I would talk to the facility and understand the rationale behind your PI's comment. @lieven.sterck gave an example of possible a scenario in their answer.
No, he is not a bioinformatician. Actually, we just did some very basic analysis with the RNAseq data using something like Rockhopper.