similar barcode warning only for 1 bp error rate

[duartemolha]

Hi ... I am using This is cutadapt 2.7 with Python 3.6.7 installed using conda

My question has to do with a warning I get when I set the error value to 0.15 (1 error in an 8 bp index allowed)
The command I use is:
cutadapt -e 0.15 --overlap 8 --no-indels -g file:indexes_anchor.fa -o selected-anchor-{name}_R1.fastq.gz -p selected-anchor-{name}_R2.fastq.gz --untrimmed-output unallocated-anchor_R1.fastq.gz --untrimmed-paired-output unallocated-anchor_R2.fastq.gz sequences_R1_001.fastq.gz sequences_R2_001.fastq.gz

I get this warning:

Building index of 48 adapters ...
WARNING: Adapters 14 'GACTAGTA' and 30 'GACTACTT' are very similar. At 1 allowed errors, the sequence 'GACTAGTT' cannot be assigned uniquely because the number of matches is 7 compared to both adapters.
Built an index containing 1192 strings.

however, if I increase the error rate to 0.25 for example (meaning I accept 2 bp errors on a 8 bp index) the error does not show again... surely if I now accept 2 bp errors instead of one I should be seeing more warnings about similarity of indexes - not less.

What am I missing?

[marcelm]

(Note that --overlap 8 is not needed since you use anchored adapters anyway. They have to occur in full.)

I cannot reproduce this at the moment. I’m testing with this command (on Cutadapt 2.7):

cutadapt --no-indels -e 0.25 -g ^GACTAGTA -g ^GACTACTT -o /dev/null /dev/null

and I do get the warning:

WARNING: Adapters 1 'GACTAGTA' and 2 'GACTACTT' are very similar. At 2 allowed errors, the sequence 'GACTAGTT' cannot be assigned uniquely because the number of matches is 7 compared to both adapters.

The check for similarity of adapters is only done when also an index is built. Do you actually get the message "Building index of 2 adapters ..." in the second case?

[duartemolha]

no I do not get the message of Building Index on the second case.

[marcelm]

So then in the second case, the index isn’t built, which is why you don’t get the warning. Did you change anything else apart from increasing the maximum allowed error rate?

[duartemolha]

No ... no change

…

On Fri, 29 May 2020, 13:00 Marcel Martin, ***@***.***> wrote: So then in the second case, the index isn’t built, which is why you don’t get the warning. Did you change anything else apart from increasing the maximum allowed error rate? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#463 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAFQIVJNCV4BSPKQKVFJGB3RT6PUZANCNFSM4NNPDD6A> .

[marcelm]

Is your indexes_anchor.fa file secret or would you be willing to share it with me? You can send it by e-mail to marcel.martin@scilifelab.se and I will not share it with anyone if that helps.

[duartemolha]

I have sent you an email... thank you for your help

[marcelm]

Thanks for the file. However, I still cannot reproduce the problem. It would be very helpful if you could run your two commands again and then copy and paste the full output up to the message "Processing reads on 1 core in paired-end mode ...".

For me, it looks like this (using /dev/null is input file names, which does not matter as the adapter index is generated before the input is read):

$ cutadapt -e 0.15 --overlap 8 --no-indels -g file:indexes_anchor.fa -o "out-{name}_R1.fastq.gz" -p "out-{name}_R2.fastq.gz" --untrimmed-output /dev/null --untrimmed-paired-output /dev/null /dev/null /dev/null
This is cutadapt 2.7 with Python 3.6.7
Command line parameters: -e 0.15 --overlap 8 --no-indels -g file:indexes_anchor.fa -o out-{name}_R1.fastq.gz -p out-{name}_R2.fastq.gz --untrimmed-output /dev/null --untrimmed-paired-output /dev/null /dev/null /dev/null
Building index of 48 adapters ...
WARNING: Adapters 14 'GACTAGTA' and 30 'GACTACTT' are very similar. At 1 allowed errors, the sequence 'GACTAGTT' cannot be assigned uniquely because the number of matches is 7 compared to both adapters.
Built an index containing 1192 strings.
Processing reads on 1 core in paired-end mode ...

Please also write here the output of conda list '(dnaio|xopen)'.