Entering edit mode
                    7.1 years ago
        gab
        
    
        ▴
    
    20
    Hi I'm using cnvkit to get CNVs from WES data. I got my .cnr right and also the .cns files with all segmentation methods except for flasso which gives me the following output:
Dropped 1 outlier bins:
  chromosome      start        end        gene     log2     depth    weight
0         14  104917061  105052263  Antitarget  12.0248  0.004364  0.066077
Dropped 26 / 207080 bins
     File "/home/gabriele/.local/bin/cnvkit.py", line 13, in <module>
    args.func(args)
  File "/home/gabriele/.local/lib/python2.7/site-packages/cnvlib/commands.py", line 629, in _cmd_segment
    processes=args.processes)
  File "/home/gabriele/.local/lib/python2.7/site-packages/cnvlib/segmentation/__init__.py", line 52, in do_segmentation
    rlibpath, rscript_path)
  File "/home/gabriele/.local/lib/python2.7/site-packages/cnvlib/segmentation/__init__.py", line 159, in _do_segmentation
    seg_out = core.call_quiet(rscript_path, '--vanilla', script_fname)
  File "/home/gabriele/.local/lib/python2.7/site-packages/cnvlib/core.py", line 36, in call_quiet
    % (' '.join(args), err))
RuntimeError: Subprocess command failed:
$ /home/gabriele/.local/bin/Rscript --vanilla /tmp/tmpH1h7vI
Segmenting 1
Segmenting 10
Segmenting 11
Segmenting 12
Segmenting 13
Segmenting 14
Segmenting 15
Segmenting 16
Segmenting 17
Segmenting 18
Segmenting 19
Segmenting 2
Segmenting 20
Segmenting 21
Segmenting 22
Segmenting 3
Segmenting 4
Segmenting 5
Segmenting 6
Segmenting 7
Segmenting 8
Segmenting 9
Segmenting X
Segmenting Y
 *** caught segfault ***
address (nil), cause 'unknown'
Traceback:
 1: .Fortran("avesmooth", ntemp, p, size, temp, result = result,     PACKAGE = "cghFLasso")
 2: plugin.NA(result, chromosome, size = smooth.size)
 3: cghFLasso(tbl$log2, FDR = 1e-04)
aborting ...
The command was
cnvkit.py segment 10P.cnr -m flasso --rscript-path /home/gabriele/.local/bin/Rscript --drop-low-coverage -o 10Pflasso.cns
Anyone have an idea about why this happens? Thanks!
Apparently Fortran code is segfaulting. cnvkit calls cghFLasso, an R package which has not been updated since June 2009, the Fortran code is probably part of cghFLasso. cnvkit includes a test for the R flasso subroutine (test/test_r.py), you could try to run it to see if cghFLasso is broken or if it is a problem with your data.