Re: [Bio-bwa-help] BWA mapping: clipping reverse reads

[Joseph F. <jos...@gm...>]

Is the problem that the 3' ends of your reverse reads (your libraries are
strand-specific?) tend not to align to the reference, *but not because of
poor base quality*?  In other words, there's some novel sequence that you
expect at the 3' ends?

In this case, would it help to run the aln command on your reverse reads
differently than for the forward reads ... say, with '-n 0.01' (so that more
differences are allowed, if I understand the behavior of the -n option
correctly)?

~Joe



On Thu, Feb 25, 2010 at 11:49 AM, Nicholas Croucher <nc...@sa...>wrote:

> Dear Heng,
>
> Sorry, I should have explained, that line was just an example from a
> small test set of reads. But the results for both of those read pairs
> are identical when I map the whole Illumina lane.
>
> Many thanks,
>
> Nick.
>
> Heng Li wrote:
> > Bwa should be able to align this read if you put this pair together
> > with, say, 100,000 other reads from the same experiment. You cannot get
> > it aligned if you just map this read. The difference is that given
> > enough reads, bwa will estimate the insert size distribution and runs
> > Smith-Waterman alignment for mates of orphan reads, which should place
> > the (54-7)bp piece of read 2, unless the alignment is really bad or the
> > insert size falls out of 3 standard error.
> >
> > Heng
> >
> > PS: I am copying to the bwa-help mailing list.
> >
> > On Thu, Feb 25, 2010 at 05:47:55PM +0000, Nicholas Croucher wrote:
> >
> >> Dear Dr. Li,
> >>
> >> I'm sorry to bother you, but I'm having some problems mapping with
> >> BWA that other users of the program I know can't help me with. I
> >> have Illumina data where the 3' end of the reverse read generally
> >> won't map correctly to the reference sequence. When I map with
> >> SSAHA, it is able to correctly trim the reads. However, with BWA, it
> >> doesn't seem to map the reverse reads. For instance, when I compare
> >> my CIGAR and SAM files, I see:
> >>
> >> 1) Cases where both reads map perfectly - which BWA copes with
> >> without any problems:
> >> SSAHA:
> >> cigar::50     IL21_4157:5:1:1321:16678/1     54 1     -     OXC141
> >> 796961     797014
> >> cigar::50     IL21_4157:5:1:1321:16678/2     1 54     +     OXC141
> >> 796855     796908
> >> BWA:
> >> IL21_4157:5:1:1321:16678    83     spn_oxc-97e10.p1k     796961
> >> 60    54M     =     796855
> >> IL21_4157:5:1:1321:16678    163   spn_oxc-97e10.p1k   796855     60
> >> 54M     =     796961
> >>
> >> 2) Cases where the distal end of the reverse read does not map
> >> properly - BWA does not map the second read.
> >> SSAHA:
> >> cigar::50     IL21_4157:5:1:1299:4085/1     54 1     -     OXC141
> >> 576304     576357
> >> cigar::50     IL21_4157:5:1:1299:4085/2     7 54     +     OXC141
> >> 576213     576260
> >> BWA:
> >> IL21_4157:5:1:1299:4085     89     spn_oxc-97e10.p1k     576304
> >> 37    54M     =     576304
> >> IL21_4157:5:1:1299:4085     181     spn_oxc-97e10.p1k     576304
> >> 0    *         =     576304
> >>
> >> I have tried altering the -q option in the 'aln' component, but
> >> without much success I'm afraid. Any help or advice would be very
> >> much appreciated.
> >>
> >> Many thanks,
> >>
> >> Nick Croucher.
> >>
>
>
> --
>  The Wellcome Trust Sanger Institute is operated by Genome Research
>  Limited, a charity registered in England with number 1021457 and a
>  company registered in England with number 2742969, whose registered
>  office is 215 Euston Road, London, NW1 2BE.
>
>
> ------------------------------------------------------------------------------
> Download Intel&#174; Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Bio-bwa-help mailing list
> Bio...@li...
> https://lists.sourceforge.net/lists/listinfo/bio-bwa-help
>



-- 
Joseph Fass
Bioinformatics Programmer
UC Davis Bioinformatics Core
joseph.fass -at- gmail.com (professional)
970.227.5928 (c)  ||  530.752.2698 (w)