[mira_talk] Re: Lots of contigs, then segmentation fault
- From: Bastien Chevreux <bach@xxxxxxxxxxxx>
- To: mira_talk@xxxxxxxxxxxxx
- Date: Wed, 4 May 2011 21:31:39 +0200
On Tuesday 03 May 2011 22:04:36 Egon Ozer wrote:
> Wow. So it wasn't just me...
>
> That's a very interesting result. The 454 reads weren't from some
> fly-by-night sequencing provider, either. Definitely one of the big
> sequencing centers that did those for me.
From what I've been told some of the 454 PE can be quite finicky and also
require a good deal of luck. I think that in your case, the sequencing
provider is very probably not at fault.
> I'd be worried that my organism
> is expected to be somewhat repeat-heavy and a fair number of those
> duplicates might be real, but the assembly numbers you were able to get
> look so good that I'm less suspicious.
Some of those duplicate reads are easily spotable by eye in an assembly:
identical start and stop, identical sequence, all in the same direection.
You'd need to have a really high coverage even in 454 to get these things in
normal sequencing.
> Of course I'll have to look at my
> own assembly and/or the results of your hybrid run to really see if it's
> believable for my bacterium...
I'll send you the link for a download by PM.
> Thanks again for all the help. Above and beyond the call of duty. If I
> have any more kids I'll consider naming them Bastien.
Uh ... you'd be actually the second person to do that. Though the first one
who did didn't do it because of MIRA :-)
> P.S. It seems if I run Mira without messing with any of the output files
> before it's finished (i.e. not opening up the log file in TextEdit to see
> how far along it is), I don't get the seg fault problem. Seems weird that
> that would be the cause of the problem, but maybe I just need to leave it
> alone and let it do its job in peace.
Cannot be the real problem, at least I would not think so.
Best,
Bastien
Other related posts:
