[mira_talk] Re: sff_extract

I'm not quite sure about your question... there is just one part1 and one part2, so how would you link part1 with part1? If you have more than two parts (meaning that there would be a sequence, a linker, a second sequence, another linker, and a third sequence, it would mean your paired-end sequences are kind of screwed... It would raise an error or a warning, AFAIK.
The reads will get names that reflect their status and you could link part1 to part2, but the linking information doesn't make it to the xml file, which I think is sound. If you have multiple linker contamination, you shouldn't really trust the pair, should you? 

Lionel


On 22 May 2009, at 18:32 , Brian Forde wrote:


Thanks again,
does having multiple linker sequence affect the paired status of the reads or will the *.part1 link with the *.part1 and part2 with part2 etc.
On Fri, May 22, 2009 at 10:42 AM, Lionel Guy <guy.lionel@xxxxxxxxx> wrote: 
Blimey, I'm stupid!
It's actually *_mlc that denotes the presence of multiple linker presence. *_pl denotes the presence of _partial_ linker (with a 10% tolerance) in a sequence. 

From the comments of the program:
'''Splits a paired end read and writes sequences into FASTA, FASTA qual 
   and XML traceinfo file. Returns the number of sequences created.
As the linker sequence may be anywhere in the read, including the ends 
   and overlapping with bad quality sequence, we need to perform some
   computing and eventually set new clip points.

   If the resulting split yields only one sequence (because linker
   was not present or overlapping with left or right clip), only one
   sequence will be written with ".fn" appended to the name.
If the read can be split, two reads will be written. The side left of the linker will be named ".r" and will be written in reverse complement into the file to conform with what approximately all assemblers expect when reading paired-end data: reads in forward direction in file. The side 
   right of the linker will be named ".f"
If SSAHA found partial linker (linker sequences < length of linker), 
   the sequences will get a "_pl" furthermore be cut back thoroughly.
If SSAHA found multiple occurences of the linker, the names will get an 
   additional "_mlc" within the name to show that there was "multiple
   linker contamination".

   For multiple or partial linker, the "good" parts of the reads are
   stored with a ".part<number>" name, additionally they will not get
   template information in the XML
   '''


Sorry for the mess...

Lionel


On 22 May 2009, at 11:24 , Brian Forde wrote:

Thanks Lionel,
While i'm on the topic I have also noticed *_mlc.part extensions as well. Can you tell me what these ones are?
On Thu, May 21, 2009 at 6:05 PM, Lionel Guy <guy.lionel@xxxxxxxxx> wrote: 
Hi Brian,
*_pl reads mean that there were several linker matches found in the read.
By the way, there has been some issues with sff_extract, what version are you using? The latest one is (AFAIK) in http://www.chevreux.org/tmp/mira_3rdparty_05-05-2009.tar.bz2 

Cheers,

Lionel


On 21 May 2009, at 18:09 , Brian Forde wrote:

Hello all,
I was having a look through the multi fasta file output of sffextract and noticed some thing. All the reads have a file extension. *.fn (normal nonpaired shotgun reads) *.f/*.r (paired reads) and *_pl.part1. What I would like to know is what are these reads the *_pl.part1? 

regards

Brian

--
"If scientists knew what they were doing they wouldn't call it research" 


--
You have received this mail because you are subscribed to the mira_talk mailing list. For information on how to subscribe or unsubscribe, please visit http://www.chevreux.org/mira_mailinglists.html 



--
"If scientists knew what they were doing they wouldn't call it research" 

============================================
Lionel Guy
Thunmansgatan 25, SE-75421 Uppsala

phone: +46 (0)18 245596
mobile: +46 (0)73 9760618
email: guy.lionel@xxxxxxxxx
============================================



--
You have received this mail because you are subscribed to the mira_talk mailing list. For information on how to subscribe or unsubscribe, please visit http://www.chevreux.org/mira_mailinglists.html 



--
"If scientists knew what they were doing they wouldn't call it research" 

============================================
Lionel Guy
Thunmansgatan 25, SE-75421 Uppsala

phone: +46 (0)18 245596
mobile: +46 (0)73 9760618
email: guy.lionel@xxxxxxxxx
============================================


--
You have received this mail because you are subscribed to the mira_talk mailing 
list. For information on how to subscribe or unsubscribe, please visit 
http://www.chevreux.org/mira_mailinglists.html

Other related posts: