[mira_talk] Re: Mappging to Reference

On Mittwoch 28 Juli 2010 Saulo Alves wrote:
> Hello Bastien,
> 
> I have used Mira 3.02 with the following parameters
> 
> ./bin/mira -project=crypto -job=mapping,genome,accurate,solexa
> COMMON_SETTINGS -GE:kcim=yes -AS:sd=yes -SK:not=8 SOLEXA_SETTINGS
> -LR:ft=fastq -CL:qc=yes -CL:qcmq=25 -GE:tismin=300:tismax=900

Do not use -CL:qc=yes -CL:qcmq=25 on Solexa data ... you're robbing yourself 
by discarding perfectly valid data, especially in some difficult to sequence 
regions. Trust -CL:pec to do the job :-)

Please also read about the -CO:msr setting in the new reference manual and the 
section "Starting a mapping assembly: paired-end data" in the new Solexa 
manual at 

  http://sourceforge.net/apps/mediawiki/mira-assembler/index.php

And please switch to 3.2.0rc1.

> Except for the late presence of the rail-reads, is there any mistake in the
> assembly I should be aware of?

See above.

> And, most important, those "deletions" in the regions not covered by any
> read, should I trust it once there are so many of them? could it be a
> problem in the sequencing instead of the program?

Simple check: use a text editor (or FASTA, the program) in your FASTQ data to 
search "by hand" some of the non-covered parts: If you find them: see comment 
on -CL:qc above :-)

B.

-- 
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: