How to handle reads with alternative adapters or primer sequences

[fxfish]

Hi,
It's a powerful tools for data analysis.
I have a question when trimming primers from paired-end reads of amplicon. The sequencing data (e.g. forward_reads) likes: AAATTTaaaaaaaaCCCGGG and GGGCCCtttttttttTTTAAA (mixed), I want all output data likes: aaaaaaaa, which means recognizing ‘GGGCCC...TTTAAA’ and reversing complementary sequence.
I must do three steps to achive this, one using -g ^AAATTT -G ^CCCGGG, and another using -g ^CCCGGG -G ^AAATTT, at last combining 1st R1.fq and 2st R2.fq into the last R1.fq.
Can I just use one step commond? -g -g or -g XXX|XXX
Thanks.

[marcelm]

This functionality doesn’t exist, yet, but I think this is actually issue #561, that is, it would be solved by implementing --revcomp for paired-end data. --revcomp already works for single-end reads, but I wasn’t sure whether it makes sense to implement it for paired-end reads.

It would work as follows. The idea is that you write -g ^AAATTT -G ^CCCGGG --revcomp. Then, for each read, this is done:

1. Normal adapter search is done as if --revcomp had not been given
2. R1 and R2 are swapped and the search is done again
3. The results from both searches are compared: If the swapped version resulted in a better match, then the swapped version is output and the unswapped version otherwise.

Note that swapping R1 and R2 is equivalent to swapping the -g and -G options.

Can you say whether that would give you the result you want?

[fxfish]

Think you very much. The function --revcomp does not work for paired-end reads. I can get the results I want by using cutadapt with several steps.

[marcelm]

I have implemented this now. Please update to Cutadapt 4.6. Then you can use --revcomp with paired-end reads.

(The release will be available on Bioconda in probably one or two hours.)

[fxfish]

wow, nice, I will update it on conda, thanks.