[mira_talk] Re: Bug?
- From: Bastien Chevreux <bach@xxxxxxxxxxxx>
- To: mira_talk@xxxxxxxxxxxxx
- Date: Wed, 3 Jun 2015 23:20:42 +0200
On 03 Jun 2015, at 22:58 , Torben Nielsen <torben@xxxxxxxxxx> wrote:
The latest one I had the issue with was around 42 M HiSeq reads (21 M pairs)
of length 99 nt.
Hrm. I have two theories.
The first one: 4.9.4 got much more aggressive read correction and junk
detection routines than earlier versions and some parameters in the late stages
count on this or else things become interesting. As you resumed an assembly
from 4.9.3, that would mean that a lot of reads have not passed these stages
and may therefore contain errors which should not be present anymore and are
therefore seen as junk.
The second theory: a couple of new routines (read correction, chimera & junk
detection) are not tuned to metagenomic data. Especially the junk detection may
be too triggerhappy there.
Could you do me a favour and start 4.9.5 fresh from start, once with the
parameters you already set and once with an additional -CL:kjd=no ? I’d be
quite interested to see the differences there.
Thanks,
Bastien
--
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:
