Vector trimming: are flanking sequences sufficient?

[sulicon]

Hi All,

I have met a strange problem when using newbler to trim vector sequence. The vector structure is:
Forward Primer ----> CDNA sequence ---->Reverse Primer

I was told only the three parts (forward & reverse primer, CDNA) could be sequenced. At first, I just used the flanking sequences, i.e. the primer sequences, for vector trimming (newbler -vt flanking.fa ...). I found some vector sequences remained in the result. For example:

vector sequence in assembled contig:
AATGCCAACTTTGTACAAAAAAaGTTGGCACC

part of the primer sequence:
AATGCCAACTTTGTACAAAAAAGTTGGCACC

I don't understand why this sequence haven't been trimmed. The contig just has one additional "a", which should be a sequencing error.

Then I added the full vector sequence into the file used for trimming. It seems that the result becomes better this time, although there are still some non-CDNA nucleotides appeared in assembled contigs ( I have aligned some of the contigs to reference sequences, and found some terminal nucleotides failed to be aligned)

I was told the flanking sequences were sufficient for vector trimming, but I observed an improved result by using full vector sequence. Is full vector sequence indeed necessary, or I have made something wrong in vector trimming?


Thanks in advance!

[flxlex]

This is not good... It's a bug (I'm pretty sure) with newbler that surprises me. You cannot change the parameters for trimming so I am afraid you are out of luck there. If using the full vector helps then use that!

One thing to try is collecting the variant sequences of your primers and using them all in your trimming file, that might help...