fastx_clipper mismatches in adapter

[fbelzile62]

I clipped an adapter sequence off a set of restriction fragments that were sequenced on an Illumina machine. My expectation was that all the sequences that had contained a full-length adapter sequence (I used a 9-bp motif) would end with a recognizable restriction site. Although the vast majority of reads do meet this expectation, to my surprise, some of the reads do not.

When I examine the sequence prior to clipping, I find that the region where a clip was performed is similar but not identical to the specified motif. In the example below, I show a case of a full-length sequence (103 bp) and the clipped sequence that was obtained.

Unclipped sequence:

CAGCAATACTGCTGGAAAGGACAGTTTCAAGANTTCCCATCCGTAATCCTTTATATAGGCGTTGGGGTGGACGCACACTAAGATGGAAAAGACCTGCTAATAT

Clipped sequence:

CAGCAATACTGCTGGAAAGGACAGTTTCAAGANTTCCCATCCGTAATCCTTTATATAGGCGTTGGGGTGGACGCACACTA

The site where the clipping occurred (AGATGGAA) is similar but not identical to the specified adapter sequence (AGATCGGAA). It seems that a missing C was tolerated and that clipping was allowed to go ahead.

I have performed the clip both in Galaxy and using the Fastx toolkit and did not see any option allowing one to accept or not mismatches in the adapter, and I got the exact same result both ways.

Any insights would be much appreciated

Francois

[maasha]

Hello,


Biopieces (www.biopieces.org) contain a couple of tools for locating and removing adaptors. Have a look at find_adaptor:

http://code.google.com/p/biopieces/wiki/find_adaptor

Cheers,


Martin

[Kennels]

Hi,

I believe it allows for max 2 mismatches in the adapter from what I gather from the source code (fastx_clipper.xml), but I could be wrong and perhaps you can also email the Galaxy user list if you haven't already done so.

cheers

[robs]

If you want to specify the number of mismatches, take a look at TagCleaner (http://tagcleaner.sourceforge.net/) that comes as a web and standalone version.