is it possible to specify how much of each read is processed by cutadapt?
I am asking because all my reads start with directly with my barcode sequence and are followed by a primer sequence.
e.g. forward read
TCAT (Barcode) CTTGGTCATTTAGAGGAAGTAA (Primer)
If i now run demux-paired with TCAT specified in my metadata as barcode sequence my resulting demultiplexed file will contain every read that has a TCAT somewhere within the read which is a bit annoying.
My current work-arround was to combine my barcode and primer sequence during demux-paired. This however won't allow any error-tolerance. Since the actual goal is to have 0-Error Tolerance in the Barcode.
What i want to achieve is pretty much, that cutadapt allows for no errors in the barcode, but allows for a defined amount of error-tolerance in the primer sequence.
e.g.
100% correct sequence
TCAT CTTGGTCATTTAGAGGAAGTAA
TCAT CTGGGTCATTTAGAGGAAGTAA
TCAT GTTGGTCATTTAGAGGAAGTAA
I use cutadapt to demultiplex based only on barcodes with -e 0. Demultiplexed paired-end files are imported into Qiime2. I trim the primers with Dada2. Dada2 will trim fixed length, irrespective of sequence.
Thank you for your reply!
I was able to fix my problem, by just accepting a lot of errors in my demux-paired output and trying to get rid of them by using the length parameter during trim-paired.