Finding all adapter occurrences in each read

• cutadapt 1.13 and Python 3.5.2
• Installed with pip Command line:

cutadapt -g file:5p_adapters.fasta --times 8 -O 15 --info-file test.tsv -o /dev/null test.fasta

I am trying to trim multiple adapters from a sequence and parse the info file to see what coordinates they were trimmed at. Unfortunately I can’t provide the actual sequences used to trim or being trimmed.

For this particular sequence I expect 3 adapters to be trimmed from the 5’ end.

I will call them 5p_a1, 5p_a2, and 5p_a3. They occur in this order in the target sequence and the adapters fasta I am using. When I run the above command however, 5p_a1 and 5p_a3 are removed but not 5p_a2.

I get the following output from cutadapt:

=== Adapter 5p_a1 ===

Sequence: AAAAAAAAAAAAAAAAAAAAAA; Type: regular 5'; Length: 22; Trimmed: 1 times.

No. of allowed errors:
0-9 bp: 0; 10-19 bp: 1; 20-22 bp: 2

Overview of removed sequences
length  count   expect  max.err error counts
8540    1       0.0     2       1

=== Adapter 5p_a2 ===

Sequence: GGGGGGGGGGGGGGGGGGGGG; Type: regular 5'; Length: 21; Trimmed: 0 times.

=== Adapter 5p_a3 ===

Sequence: CCCCCCCCCCCCCCCCCCCCC; Type: regular 5'; Length: 21; Trimmed: 1 times.

No. of allowed errors:
0-9 bp: 0; 10-19 bp: 1; 20-21 bp: 2

Overview of removed sequences
length  count   expect  max.err error counts
113     1       0.0     2       1

So 5p_a2 not trimmed. However if I remove 5p_a3 from the adpaters file I get the following output:

=== Adapter 5p_a1 ===

Sequence: AAAAAAAAAAAAAAAAAAAAAA; Type: regular 5'; Length: 22; Trimmed: 1 times.

No. of allowed errors:
0-9 bp: 0; 10-19 bp: 1; 20-22 bp: 2

Overview of removed sequences
length  count   expect  max.err error counts
8540    1       0.0     2       1

=== Adapter 5p_a2 ===

Sequence: GGGGGGGGGGGGGGGGGGGGG; Type: regular 5'; Length: 21; Trimmed: 0 times.
No. of allowed errors:
0-9 bp: 0; 10-19 bp: 1; 20-21 bp: 2

Overview of removed sequences
length  count   expect  max.err error counts
60      1       0.0     2       0 1

5p_a3 trimmed!

I can see that the error counts are slightly different for 5p_a2 - 0 1 compared to 1. But both have 1 total error so I am confused as to why this causes 5p_a2 to be skipped.

[awgymer]

Ah ok, I think I understand what your query is now.

So the assumption is that I don’t actually know the order the adapters occur in, or at least I may not. Part of the reason for performing this step is that, assuming we expected to see A B C in the A C B C scenario we want to know that actually C could also trim there. So the intention would be that it would detect as many as possible by default I suppose , with --times in this case acting as a sort of limiter to enhance speed.

In reality, I know roughly how many times I expect the adapter sequences to occur (once per adapter) but not how many of the sites are actually present (i.e. 2 out of 4 adapters v all 4).

I will take some time over the holidays to play with some sort of Modifier class then, and see if I can come up with something which seems like it might be useful to more than just me.

[marcelm]

I’m closing this now as you suggested, but I’m adding it to my long-term to-do list (whatever that means).