Cutadpapt for paired-end sequencing reads

Question

[flsnike]

Hi dear All, I have been recently doing rna-seq analysis. In the fastqc report it shows overpresented sequcences(1.7%) but with "NO Hit" in the "Possible Source" column. DO I need to trim them off? As it is paired-end reads, how to present the "adapter" when using cutadapt ?

For instance, the sequence I wanna trim is "GTGCTCTT", so the command should be like

cutadapt -a GTGCTCTT -A AAGAGCAC -o out.1.fastq -p out.2.fastq reads.1.fastq reads.2.fastq

where the sequence following "-A" is reverse-complementary to the sequence behind "-a"

Can anyone tell me if it is right to write command above ? or if not correct, plz show me the correct command

Thanks in advance!!

Answer 1

[Brian Bushnell]

Those are too short for confident adapter identification. For paired-end reads, the best approach is to use BBDuk as illustrated at the top of this thread. This will use all commonly-used Illumina adapter-sequences for trimming as well as overlap-detection to trim adapter fragments too short to catch via sequence-matching. If you believe that your data contains adapters not present in BBMap's adapters.fa file, you can determine the sequence yourself like this:

bbmerge.sh in=r1.fq in2=r2.fq outa=adapters.fa reads=2m

The sequences you mention ARE part of normal TruSeq adapter sequences, by the way - just not at the beginning. So, in addition to being too short, they are probably not the correct sequences to use for good adapter trimming.