Is It Necessary To Do Read Trimming For Rna-Seq Data?

Question

[Junfeng]

Hi, we obtained pared-end RNA-Seq data from Illumina.Unfortunately, the reads are not of the same length. The length of first reads is 100bp, while the second reads is 97bp. I have two questions about the read length.

(1). I wonder whether it is necessary to do read trimming so that both reads are of the same length. I know that most mapping tools are not care the disparate read lengths between a pair.

(2). In addition, I wonder whether it is necessary to use trimming tools to cut a read if below a threshold quality. I think most mapping tools can deal with base call qualities, so it is no need to trim reads. Is it correct? Thanks.

Answer 1

[Vitis]

Usually I wouldn't do trimming since as you said, the mapping tools would take care of it. But if I'd like to call SNPs with some confidence, I'll trim the bad bases. For me the threshold was determined empirically by plotting the qualities.

Answer 2

[Ryan Thompson]

There's no need for the reads to be of identical lengths. As for quality-trimming, I would say it is recommended if you are doing SNP calls, but not otherwise.

However, regardless, I would recommend using SeqPrep on your data before anything else. It can detect when the insert size is shorter than the read length and trim accordingly.

[Junfeng]

Hi Vitis and Ryan, thanks for your kind reply. I will trim reads according to your suggestions.