This is a re-posting of an older topic that I have still not resolved. This was covered in Error Running Breakdancer-Max: Max Kahan Error: 0 Breakdancer Installation/Running Problems: Error/Suspicious Messages During Installation prior threads. In the two prior threads it was suggested that I try updating to breakdancer version 1.4.4. I did this, and it did not resolve the issue. I'm still getting the same error.
Old post, updated for latest version follows:
I'm attempting to run breakdancer-max, and am running into an error that I can not find documentation on. My script looks like this:
=======
#breakdancer directory
BD="/share/apps/ngs-ccts/breakdancer-1.4.4"
#File name
FILE="CNTL3487.sort.bam"
#Output directory
DIR="breakdancer_test"
#Create config file
perl $BD/perl/bam2cfg.pl $FILE > $DIR/CNTL3487.cfg
$BD/bin/breakdancer-max -d CNTL3487 $DIR/CNTL3487.cfg > $DIR/CNTL3487.ctx
======
And the output that I get looks like this:
======
[Tue Dec 10 13:32:15 2013 /share/apps/ngs-ccts/breakdancer-1.4.4/perl/bam2cfg.pl] Processing bam: CNTL3487.sort.bam
[Tue Dec 10 13:32:18 2013 /share/apps/ngs-ccts/breakdancer-1.4.4/perl/bam2cfg.pl] selected_libs is : 1
[Tue Dec 10 13:32:18 2013 /share/apps/ngs-ccts/breakdancer-1.4.4/perl/bam2cfg.pl] Closing BAM file
[Tue Dec 10 13:32:18 2013 /share/apps/ngs-ccts/breakdancer-1.4.4/perl/bam2cfg.pl] Send TERM signal for 3124
[Tue Dec 10 13:32:20 2013 /share/apps/ngs-ccts/breakdancer-1.4.4/perl/bam2cfg.pl] samtools pid process 3124 is still there...
[Tue Dec 10 13:32:20 2013 /share/apps/ngs-ccts/breakdancer-1.4.4/perl/bam2cfg.pl] invoking kill -9 on 3124 ...
[Tue Dec 10 13:32:20 2013 /share/apps/ngs-ccts/breakdancer-1.4.4/perl/bam2cfg.pl] Closing samtools process : 3124
Max Kahan error: 0
=====
All of the output except for the Max Kahan error : 0 line is generated by bam2cfg.pl, and the Kahan line is generated by breakdancer-max. This also generates two files in the home directory (same directory I'm running in, and the one containing CNTL3487.sort.bam) with a size of 0 called CNTL3487.NA.2.fastq and CNTL3487.NA.2.fastq. It also creates a CNTL3487.ctx file in the breakdancer_test/ folder that looks like this:
======
#Software: 1.4.4 (commit ad4dca4)
#Command: /share/apps/ngs-ccts/breakdancer-1.4.4/bin/breakdancer-max -d CNTL3487 breakdancer_test/CNTL3487.cfg
#Library Statistics:
#CNTL3487.sort.bam mean:605.16 std:159.56 uppercutoff:1222.92 lowercutoff:0 readlen:144 library:NA reflen:3005868224
seqcov:0.238852 phycov:0.501888 1:100 2:650 4:636 8:50 32:2722
#Chr1 Pos1 Orientation1 Chr2 Pos2 Orientation2 Type Size Score num_Reads num_Reads_lib CNTL3487.sort.bam
======
Note that the .bam file used here contains sequence reads from genomic material that first went through a capture step to isolate a region from Chr1.
I'm sorry to hear that 1.4.4 did not do the trick. Based on your earlier questions, I think I'm stumped without a concrete example to debug with. Is it at all possible for you to share a subset of these data where you believe there should be a prediction from IGV?
The file in the example may or may not have a novel breakpoint. I have another .bam file in which there definitely should be a breakpoint detected. Now, as for IGV, the breakpoint in question is in a very difficult part of the genome (lots of repetitive sequence), so it won't be clearly visible in IGV.
How can I get you the .bam file? I'm not seeing an attachment option here.
Hard stuff! I ran several times breakdance-max and never had a problem. You might try to set -r 1 (require one read to support an event istead of two) Also. If it's a capture experiment, maybe you have very high coverage. -x controls the maximum control of a region, and by default will remove all regions with coverage greater than 1000. Just guessing...
Solved see here: Breakdancer Max Kahan Error:0