[mira_talk] Re: Problems using sff extract
- From: "Bastien Chevreux" <bach@xxxxxxxxxxxx>
- To: mira_talk@xxxxxxxxxxxxx
- Date: Thu, 16 Dec 2010 09:02:18 +0100 (MET)
>From: Ganga Jeena
> Thanks it worked with 0.2.9
0.2.9 and 0.2.9 are identical except I re-activated a check for ssaha2
runability from Python. If 0.2.9 worked, 0.2.8 should have too.
> Here for only 80709 paired end reads sequences generated could have been
> maximum
> double 161418 but they are 7 times 754371 ( Converted 441320 reads into
> 754371 sequences.)
> How is it possible ?
> How exactly does the sff_extarct works?
SSAHA2 for finding linker sequences, splitting reads at these places. For
details, please read the function description comment in the function
split_paired_end(data, sff_fh, seq_fh, qual_fh, xml_fh):
of sff_extract
> Does it not only take sequences with linker and discard the others which
> either had no liker
> or had linker in far-end which when separated could not have the other pair
> end?
No, why throw away data which could still be useful?
> Is the .r for reverse and .f for forward strand of same sequence ends ?
Yes.
> What does this .fn indicate??
> Why are nnn appended to end of the sequences ?
See function comments pointed to above-
> Why Most of the sequences are in small letters ?
Clipped sequences in the SFF, see Roche documentation.
B.
Other related posts:
