[EMBOSS] Is vectorstrip gapless by design or is it a bug ?

[Charles Plessy]

Le Mon, Feb 26, 2007 at 03:15:15PM -0000, pmr at ebi.ac.uk a écrit :
> Dear Charles,
> 
> > In the following example, vectorstrip identifies the first primer with six
> > mismatches, although it has only two. It means that if I run vectorstrip
> > with
> > a -mismatch value lower that 29, I do miss the primer.
> 
> vectorstrip is indeed gapless by design. The algorithm is rather crude and
> could be updated. I am currently looking into other vectorstrip issues and
> now is a good time to ask questions about it.
> 
> Being gapless, you have to look at the number of mismatches without
> inserting gaps. I believe it was designed with the asusmption that 5'
> vector matches would be in good quality sequence.

Dear Peter and Jon,

thank you for your answers.

The manual page of vectorstrip says that it "is suitable for use with
low quality sequence data", so it would be definitely good to either
update the algorithm or the documentation.

The way I am using vectorscript is to look for PCR primers in sequence
reads from PCR products cloned in TA vectors. In that case, as there is
no directionality, I have to write two entries in the vector files.
Also, sometimes there are PCR artifacts amplified with one primer only.
So maybe an option to deal with this simply could be useful if I am not
the only one to use vectorstrip for this?

Have a nice day,

-- 
Charles Plessy
http://charles.plessy.org
Wako, Saitama, Japan