Re: [wgs-assembler-users] ContigContainment failed

[Walenz, B. <bw...@jc...>]

Hi, Ole-

After merging two scaffolds, we do a least squares estimate of the gap sizes
in the new scaffold.  If those gap sizes imply two contigs should be merged
(via a negative gap) we try to merge them.  This merge failed for reasons I
haven't looked into.  Probably bad sequence alignment.

I think you can safely disable this assert.  Gap size estimation might run
through a few more iterations (with the same result) and eventually give up.
The scaffold will have (slightly?) bogus gap sizes.  There are,
unfortunately, plenty of other scaffolds that fail to get gap size
estimates, so this isn't a disaster.

b


On 6/11/12 1:51 PM, "Ole Kristian Tørresen" <o.k...@bi...> wrote:

> Hi,
> I got an assertion fail while scaffolding today. Error message:
> CreateNewGraphNode()-- Contig 16085537
> * Create a contig 16085537 in scaffold 131913
>>>> Fixing up suspicious overlap (15370134,16085537,I) (ahg:-340 bhg:-143) to
>>>> (15370134,16085537,O) (ahg:143 bhg:340) len: 137
> * FOEXS: SUSPICIOUS Overlap found! Looked for
> (16085537,15370134,I)[20,1044] found (15370134,16085537,O) 137
> WARNING:  InsertChunkOverlap()-- Chunk overlap already exists.
> NEW 15370134,16085537,O - min/max 131/142 0/0 erate 0.100000 flags
> 10000 overlap 137 hang 0,0 qual 0.000000 offset 0,0
> OLD 15370134,16085537,O - min/max 20/1044 20/1044 erate 0.100000 flags
> 10001 overlap 137 hang 143,340 qual 0.000000 offset 0,0
> WARNING:  CreateChunkOverlapFromEdge()-- Chunk overlap already exists.
>  Keeping old overlap.
> NEW 15370134,16085537,O - min/max 131/142 0/0 erate 0.100000 flags
> 10000 overlap 137 hang 0,0 qual 0.000000 offset 0,0
> OLD 15370134,16085537,O - min/max 20/1044 20/1044 erate 0.100000 flags
> 10001 overlap 137 hang 143,340 qual 0.000000 offset 0,0
> * Switched right-left, orientation went from I to O
> * CreateAContigInScaffold() failed.
> ContigContainment failed.
> cgw: LeastSquaresGaps_CGW.C:1410: RecomputeOffsetsStatus
> RecomputeOffsetsInScaffold(ScaffoldGraphT*, CDS_CID_t, int, int, int):
> Assertion `0' failed.
> 
> 
> This is about 30x coverage with Illumina reads; 10x combined reads
> (180 bp insert, 100 nt reads and combined with FLASH), 10x PE 100 nt
> reads 300 insert and 5k mate pair 100 nt reads, all error corrected
> with Quake. It's probably not an optimal combination, but I'm testing
> a bit and interested in the result.
> 
> Using bogart and dnc, the 5k library compared against the other two.
> Other options are default.
> 
> Is that assertion something that can be fixed?
> 
> Thank you.
> 
> Ole
> 
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> wgs-assembler-users mailing list
> wgs...@li...
> https://lists.sourceforge.net/lists/listinfo/wgs-assembler-users