Is It Ok To Use One End Of A Set Of Paired-End Reads As A Set Of Single Reads?

Question

[Tianyang Li]

Hi,

I'm wondering if it's OK to use one end of a set of paired-end reads as a set of single reads?

I just want to get a single read data set of a sample, but currently there's only paired-end data available. Mapping isn't that crucial to me right now.

I don't have a very in depth knowledge of the sequencing process, but it makes sense to me.

Thanks!

[Mitchell]

Please be more context specific. Current answers assume you are performing mapping.

[Tianyang Li]

I've edited it, I'm more interested in generating a single read data set.

Answer 1

[Jorge Amigo]

of course it is OK, if with OK you mean that you aim to obtain single end results. the idea of a paired-end experiment is basically to improve the quality of the mapping process, hence you'll be more confident on your BAM alignments specially on difficult regions to sequence. after all, the paired-end reads alignment is no other than mapping each set independently, and then perform a pairing between all those results removing from final BAM all previous reads that don't match the pairing requirements. it's like a quality control step which you may avoid if desired, although you'll be lowering the power of the experiment.

[Khader Shameer]

+1. "The idea of a paired-end experiment is basically to improve the quality of the mapping process". Some folks just don't get this idea :) !.

Answer 2

[Conan]

IMO maybe it's feasible to map single ends data for SNP calling. But for SV calling or the data is RNA-Seq you'd better mapping reads with pairs.

[Jorge Amigo]

that's a generally spread thought which I don't share. good SNP calling depends on good mapping quality as much as SV calling, but as we are so used in research to have so many false positives we tend to feel fine with all those filters we have to apply afterwards. the more effort you do to increase the mapping quality, the best results you'll obtain when calling variants. it's the trade-off you have to continuously balance through all your ultrasequencing experiments.

Answer 3

[Mitchell]

Why would you want to do this? If you're assembling best to use paried-end.

[Tianyang Li]

I'm trying to verify a method for RNA-Seq analysis, but the amount of data is limited.

[Obi Griffith]

Hey Mitchell. Welcome to biostar. Just a little reminder for future. These kind of question/comments are better added as a 'comment' to the original question or another answer. Not as an 'answer' itself.