"Histogram of adapter lengths" reports confusing numbers about cutadapt HOT 4 CLOSED

From [email protected] on February 15, 2011 12:31:11

Finding an adapter of length 33 in a read of length 30 is actually possible due to deletions in the read. For example, if your adapter is ACGTACGT (8 nt) and the read is AGTACGT (7 nt), then an alignment could be this:
adapter: ACGTACGT
read: A-GTACGT
In this example, cutadapt would report that it has trimmed an adapter of length 8 although the read has only length 7.

The default error rate of cutadapt allows 10% errors. In your case (reads of length 30), that amounts to at most 3 errors, which is why you see at most a removed adapter length of 33 (read length 30 plus 3 deletions).

I guess it makes sense to change the histogram such that more intuitive numbers are reported. I'll keep your report open until I have done so. Thanks for your report!

from cutadapt.

From [email protected] on May 19, 2011 14:24:45

Hi there,

When I see:

Histogram of adapter lengths
length count
3 113604

What the length 3 represents for me (consider the same parameters and data as was explained above)?.

Thanks.

from cutadapt.

From [email protected] on May 20, 2011 08:05:10

This means that 113604 of the input reads end with 'CCC' (the beginning of the given adapter), and that in all those cases, the 'CCC' was trimmed from the read.

from cutadapt.

From [email protected] on July 09, 2012 15:20:05

This has been changed in cutadapt 1.1: The histogram now reports the lengths of trimmed sequences.

Status: Fixed

from cutadapt.