Re: [Bio-bwa-help] fail to infer insert size

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 454-5900

The patch is the one I sent on June 22nd to the list.

-Chris

On Jun 23, 2011, at 1:37 AM, Fantine Mordelet wrote:

> Many thanks for these precisions!
> Indeed, I checked on a run which completed ok and the percentage of properly paired reads is around 1%
> So the patch you mention, is it the one you sent on the mailing list yesterday?
> 
> Fantine
> 
> 2011/6/23 Chris Berthiaume <chrisbee@u.washington.edu>
> Hi Fantine,
> 
> A note on SOLiD paired-end data, not necessarily a solution to your crash though:
> BWA's sampe doesn't fully support SOLiD paired-end data.  sampe expects that SOLiD paired data will have reads oriented same strand same direction.
> 
> R3 ------>                   F3 ------>
> 
> Whereas paired-end data will be same strand, opposite direction
> 
> F3 ------>                    <------ F5
> 
> sampe will happily process SOLiD paired-end data, but most of your pairs will not be considered proper because they don't have the expected orientation.  What you will see is a very low percentage of your pairs are called "proper" in your BAM file.
> 
> samtools view -f2 a.bam | wc -l  # number of properly paired reads
> samtools view -f1 a.bam | wc -l  # number of paired reads
> 
> For the parts that do complete OK, what's your percentage of properly paired read pairs?  For example, for one of my SOLiD PE data sets when I use bwa sampe v 0.5.9 I see <1% of read pairs flagged as properly paired, but if I apply a patch to correct the orientation this becomes 94%.
> 
> -Chris
> 
> --
> Chris Berthiaume
> Center for Environmental Genomics
> University of Washington
> 
> On Jun 22, 2011, at 8:36 AM, Fantine Mordelet wrote:
> 
> > ok I will try that.
> > Something strange too is that it fails so close from the end (after having processed 1835008 sequences over 2000000!). Do you think it might be because the last "bunch" of reads is not big enough?
> >
> > Fantine
> >
> >
> > 2011/6/22 Tom Blackwell <tb...@um...>
> >
> > One might experiment with supplying the "-a" parameter to bwa sampe. Using "-a 4500" might be a good choice to start.  See the man page.
> >
> >                                                -  tom blackwell  -
> >
> >
> > On Wed, 22 Jun 2011, Fantine Mordelet wrote:
> >
> > Hello,
> >
> > Thanks for your answser!
> > However, I ran Bwa in parallel on several sets of 2000000 read pairs. And
> > only a few of them yielded this error. Also, on previous SOLiD data (where I
> > had much fewer reads), I did not have this problem and the pairing worked
> > perfectly fine...
> >
> > Fantine
> >
> > 2011/6/22 Brent Pedersen <bpe...@gm...>
> >
> > 2011/6/22 Fantine Mordelet <fan...@en...>:
> > Dear Bwa users,
> >
> > I have a question concerning the sampe step of Bwa aligner.
> > I tried to map 2 000 000 SOLiD paired-end reads and I get the following
> > warning in the log:
> >
> > [bwa_sai2sam_pe_core] 1835008 sequences have been processed.
> > [bwa_sai2sam_pe_core] convert to sequence coordinate...
> > [infer_isize] fail to infer insert size: too few good pairs
> >
> > My problem is that it seems to prevent Bwa from processing the remaining
> > sequences. Is that normal? The sam file generated contains errors and is
> > not
> > complete (most of times, samtools fails to transform it to a bam file).
> > Besides the job error output indicates a segmentation fault.
> >
> > I am using version 0.5.9
> >
> > Did anyone come across the same problem?
> > Any suggestion or help would be greatly appreciated.
> >
> > Cheers,
> > --
> > Fantine Mordelet, PhD
> > Statistical machine learning and modelling of biological systems
> > Inserm u900/Institut Curie/Mines ParisTech
> > 26 rue d'Ulm
> > 75005 Paris - France
> >
> > Email : fan...@mi...
> > Tél : 01 56 24 69 88
> >
> >
> > ------------------------------------------------------------------------------
> > Simplify data backup and recovery for your virtual environment with
> > vRanger.
> > Installation's a snap, and flexible recovery options mean your data is
> > safe,
> > secure and there when you need it. Data protection magic?
> > Nope - It's vRanger. Get your free trial download today.
> > http://p.sf.net/sfu/quest-sfdev2dev
> > _______________________________________________
> > Bio-bwa-help mailing list
> > Bio...@li...
> > https://lists.sourceforge.net/lists/listinfo/bio-bwa-help
> >
> >
> >
> > BWA expects mate-pairs for solid, not paired-end.
> >
> >
> >
> >
> > --
> > Fantine Mordelet, PhD student
> > Statistical machine learning and modelling of biological systems
> > Inserm u900/Institut Curie/Mines ParisTech
> > 26 rue d'Ulm
> > 75005 Paris - France
> >
> > Email : fan...@mi...
> > Tél : 01 56 24 69 88
> >
> >
> >
> > --
> > Fantine Mordelet, PhD student
> > Statistical machine learning and modelling of biological systems
> > Inserm u900/Institut Curie/Mines ParisTech
> > 26 rue d'Ulm
> > 75005 Paris - France
> >
> > Email : fan...@mi...
> > Tél : 01 56 24 69 88
> > ------------------------------------------------------------------------------
> > Simplify data backup and recovery for your virtual environment with vRanger.
> > Installation's a snap, and flexible recovery options mean your data is safe,
> > secure and there when you need it. Data protection magic?
> > Nope - It's vRanger. Get your free trial download today.
> > http://p.sf.net/sfu/quest-sfdev2dev_______________________________________________
> > Bio-bwa-help mailing list
> > Bio...@li...
> > https://lists.sourceforge.net/lists/listinfo/bio-bwa-help
> 
> 
> 
> 
> 
> 
> 
> -- 
> Fantine Mordelet, PhD student
> Statistical machine learning and modelling of biological systems
> Inserm u900/Institut Curie/Mines ParisTech
> 26 rue d'Ulm
> 75005 Paris - France
> 
> Email : fan...@mi...
> Tél : 01 56 24 69 88