[szilva]

Ditto, it is a pain also for us, would be very nice to have it fixed in GATK. They say in http://getsatisfaction.com/gsa/topic...ent_i_d_events that this is a buggy CIGAR but I doubt - there is nothing to say adjacent I/Ds are forbidden. New aligners that are not based on Smith-Waterman or BW can generate valid alignments with CIGARs like this. Should we ask them again on the GATK forum to have an other look?

[siwright]

Stampy->GATK

Yes, I was using Stampy 0.1.12 to generate the alignments. The problem is fixed in Stampy 0.1.13 as far as I can tell. I've generally been pretty happy with the indels generated from a Stampy > realign > score recal > unifiedgenotyper pipeline.

Ah, OK excellent. Thanks so much, I'll give version 13 a try. Yes, I really want to be able to maintain use of stampy as the aligner- it does an excellent job with highly polymorphic alignments.

[Gavin_Sherlock]

Exact same problem

I'm using version 13 of Stampy, but have exactly the same problem when trying to call indels with -glm INDEL, namely:

"SAM/BAM file SAMFileReader{/Volumes/Data/sherlock/Pseudomonas/All_PAO1_strains.realigned.recal.bam} is malformed: Adjacent I/D events in read MAGNUM:2:65:15227:4196#0"

The reads in question have the following alignments:

MAGNUM:2:65:15227:4196#0 83 NC_002516.2 33359 5 4M4D2M11I19M = 33371 -16 TGAACCGCTCTTCTGATCTCGGCGGTGTATTCGCCG =8?8=C7HAF?AF<;2=G@G:CH:3?/,4:8A1;3: RG:Z:GSRG000011 NM:i:1 MQ:i:5 OQ:Z:=7?5A@4A@BA@B:@@7D@D@BA@BB338:86868B PQ:i:431 UQ:i:23 XQ:i:228
MAGNUM:2:65:15227:4196#0 163 NC_002516.2 33371 5 18M11I6D7M = 33359 16 CGGCGGTGTATTCGCCGAGATCGGAAGAGCGTCGTG @FGEFIBKB?DBHJGHKCID=FJJ9FB9K0KCIKCB RG:Z:GSRG000011 NM:i:1 SM:i:0 MQ:i:5 OQ:Z:IHIGFGIIIDGGGEEIIBDI@DGG8G?DG2GDGGGD PQ:i:431 UQ:i:42 XQ:i:204

The file worked fine for calling SNPs. Does anyone know a fix or workaround? Can one simply:

samtools view All_PAO1_strains.realigned.recal.bam | grep -v "MAGNUM:2:65:15227:4196#0" > fixed.sam

add in the header, index and rerun? I guess I'll try.

[NGSfan]

Hi Gavin,

try adding -rf BadCigar and see what happens.

Let me know if it works for you

[Gavin_Sherlock]

Hi Gavin,

try adding -rf BadCigar and see what happens.

Let me know if it works for you

I've set it running, and it has passed the point of failure, so potentially looks good. If it fails elsewhere, I'll follow up, otherwise assume that it worked.

Many thanks for the help.

[dustinlong]

Same problem--running identical pipelines on some exome data with several aligners and only got this error with the Stampy (v14) BAMs (specifically for -INDEL or -BOTH, but not with -SNPS). Will try the above--Thanks

[MQ-BCBB]

Same problem--running identical pipelines on some exome data with several aligners and only got this error with the Stampy (v14) BAMs (specifically for -INDEL or -BOTH, but not with -SNPS). Will try the above--Thanks

With version GenomeAnalysisTK-1.5-32, I still have the same problem. Using the -rf BadCigar allows the UnifiedGenotyper analysis to proceed.