Does cutadapt trim trailing N's first and then use max_n to filter reads?

Question

Background I want to trim leading and (likely just) trailing N's from my WES (Illumina NextSeq500) reads with cutadapt (--trim-n).

I also want to filter for reads with > 40%? N's (--max-n).

Question Does cutadapt first remove the trailing/leading N's before removing reads with >X% N's? This is because if, hypothetically, 40% of the N's are on the end of the reads, I want to keep them.

Cheers and thanks in advance.

Answer 1

Yes, it behaves just like you expect, where a read with >40% Ns but only at the ends will still be kept (after trimming).

---

Cutadapt runs in two stages, with read modification first and read filtering after. So first, --trim-n removes Ns at the start and end of reads are removed, and only then is a rule like --max-n applied.

Here's an example where the first 5 sequences have Ns at the ends and the next 5 have Ns interspersed:

>seq01  0%
CACATCAAGAGGTCCTTGGCGACTATTAGCGCCCGGCGGACCCAATCGTA
>seq02 20%
NNNNNCAAGAGGTCCTTGGCGACTATTAGCGCCCGGCGGACCCAANNNNN
>seq03 40%
NNNNNNNNNNGGTCCTTGGCGACTATTAGCGCCCGGCGGANNNNNNNNNN
>seq04 60%
NNNNNNNNNNNNNNNTTGGCGACTATTAGCGCCCGNNNNNNNNNNNNNNN
>seq05 80%
NNNNNNNNNNNNNNNNNNNNGACTATTAGCNNNNNNNNNNNNNNNNNNNN
>seq06 10%
CACATCANGAGGTCCTTNGCGACTATTNGCGCCCGGCNGACCCAATCNTA
>seq07 20%
CANATCANGAGGNCCTTNGCGANTATTNGCGCNCGGCNGACCNAATCNTA
>seq08 40%
CNNATCNNGAGNNCCTNNGCGNNTATNNGCGNNCGGNNGACNNAATNNTA
>seq09 60%
CNNANCNNGNGNNCNTNNGNGNNTNTNNGNGNNCNGNNGNCNNANTNNNA
>seq10 80%
CNNNNNNNGNNNNCNNNNGNNNNTNNNNGNNNNCNNNNGNNNNANNNNNA

With just read filtering:

$ cutadapt --quiet --max-n 0.4 seqs.fa -o -
>seq01  0%
CACATCAAGAGGTCCTTGGCGACTATTAGCGCCCGGCGGACCCAATCGTA
>seq02 20%
NNNNNCAAGAGGTCCTTGGCGACTATTAGCGCCCGGCGGACCCAANNNNN
>seq03 40%
NNNNNNNNNNGGTCCTTGGCGACTATTAGCGCCCGGCGGANNNNNNNNNN
>seq06 10%
CACATCANGAGGTCCTTNGCGACTATTNGCGCCCGGCNGACCCAATCNTA
>seq07 20%
CANATCANGAGGNCCTTNGCGANTATTNGCGCNCGGCNGACCNAATCNTA
>seq08 40%
CNNATCNNGAGNNCCTNNGCGNNTATNNGCGNNCGGNNGACNNAATNNTA

With trimming also:

$ cutadapt --quiet --max-n 0.4 --trim-n seqs.fa -o -
>seq01  0%
CACATCAAGAGGTCCTTGGCGACTATTAGCGCCCGGCGGACCCAATCGTA
>seq02 20%
CAAGAGGTCCTTGGCGACTATTAGCGCCCGGCGGACCCAA
>seq03 40%
GGTCCTTGGCGACTATTAGCGCCCGGCGGA
>seq04 60%
TTGGCGACTATTAGCGCCCG
>seq05 80%
GACTATTAGC
>seq06 10%
CACATCANGAGGTCCTTNGCGACTATTNGCGCCCGGCNGACCCAATCNTA
>seq07 20%
CANATCANGAGGNCCTTNGCGANTATTNGCGCNCGGCNGACCNAATCNTA
>seq08 40%
CNNATCNNGAGNNCCTNNGCGNNTATNNGCGNNCGGNNGACNNAATNNTA