Breakdancer-1.3.6 - Why Are All Svs Reported As Coming From Normal?

Question

[hcarte10]

I am posting at the request of a friend at WashU. I am not sure if this issue has been resolved with the latest release. I was following the pipeline described here: http://breakdancer.sourceforge.net/pipeline.html

Using BreakDancer-1.3.6 (at least that's what the readme file says) with the following command lines:

perl bam2cfg.pl -g -h Tumor.bam Normal.bam > breakdancer.cfg
breakdancer-max breakdancer.cfg > breakdancer_SVpred.out

In the output file, it looks like all putative SVs are reported as coming from Normal.bam (column 11), even when I switch the order in the first step. Am I reading the output file incorrectly?
Thanks!
-HC

My .cfg file looks like this:

readgroup:NA    platform:illumina       map:Tumor.bam      readlen:100.00  lib:NA  num:10001       lower:69.53     upper:606.49    mean:322.73     std:67.07       SWnormality:-17.32      flag:0  exe:samtools view
readgroup:NA    platform:illumina       map:Normal.bam      readlen:100.00  lib:NA  num:10001       lower:60.58     upper:693.63    mean:366.08     std:79.10       SWnormality:-11.69      flag:0  exe:samtools view

My output file looks like:

 #Software: 1.3.7-unstable-6-071043d
   #Command: bin/breakdancer-max breakdancer.cfg
   #Library Statistics:
   NA does not contain any normals
   #Normal.bam mean:366.08     std:79.1        uppercutoff:693.63      lowercutoff:60.58       readlen:100     library:NA      reflen:2951015189       seqcov:0.604493 phycov:1.10646  1:3323  2:19395 3:42    4:9847  8:3523  32:943645<br>
   #Chr1   Pos1    Orientation1    Chr2    Pos2    Orientation2    Type    Size    Score   num_Reads       num_Reads_lib   Normal.bam     Tumor.bam<br>
chr1    3418549 12+8-   chr1    3418616 12+8-   ITX     -302    99      3       Normal.bam|3        NA      NA<br>
chr1    142537742       3+12-   chr1    142541128       31+9-   ITX     2706    99      2       Normal.bam|2        1191376.25      NA<br>
chr1    210827199       31+59-  chr1    210827331       31+59-  ITX     -297    99      7       Normal.bam|7        NA      NA<br>
chr2    89872705        14+109- chr2    89872940        14+109- INV     -365    99      2       Normal.bam|2        NA      NA<br>
chr2    133017381       746+65- chr2    133018487       5+124-  DEL     1226    99      31      Normal.bam|31       231473.78       NA<br>
...<br>

Answer 1

[ernfrid]

I believe the problem is that the readgroup and library ids must be globally unique. Since they are both NA for each file, BreakDancer cannot differentiate between the two samples. To fix this, you would need to add readgroups with different ids to each file. This can be done with Picard's AddOrReplaceReadGroups command.

[Chirag Nepal]

Could you please tell how to extract information to pass on AddOrReplaceReadGroups, either from mapped BAM or Fastq files.