[BioC] error in exomeCopy - IRanges?

Michael Love love at molgen.mpg.de
Tue Aug 7 16:51:31 CEST 2012

   On 08/07/12 15:41, Lescai, Francesco wrote:

   Thanks Mike,

   that seems exactly the case.

   Does it mean I have to re-run both the reads counting and computation?

   or can I fix it somehow?



   On 6 Aug 2012, at 23:46, Mike Love <[1]love at molgen.mpg.de> wrote:

     On Tue, Aug 7, 2012 at 12:11 AM, Lescai,
     Francesco <[2]f.lescai at ucl.ac.uk> wrote:

     Hi there,
     I'm having an error in exomeCopy when I count the CNVs.
     No previous error in any of the other steps illustrated in the vignette
     the computation of the fit list didn't give any warning and it seems to be
     formatted correctly
     > fit.list[[1]][1]
     ExomeCopy object
     sample name: UCLG_502_ATCACG_L003
     percent normal state: 79.87%
     > summary(fit.list$UCLG_502_ATCACG_L003$`1`)
        Length     Class      Mode
             1 ExomeCopy        S4
     but then I get
     > compiled.segments <- compileCopyCountSegments(fit.list)
     Error in .Call2("solve_user_SEW0", start, end, width, PACKAGE = "IRanges")
       solving row 1: negative widths are not allowed

   Hi Francesco,
   I have seen this error before when the target GRanges has duplicate or
   unsorted regions, and fixed this in the devel branch.  You could try to
   reduce the target GRanges first, before counting sample read counts.  Could
   you try:
   all.equal(target, reduce(target))
   If these are in fact equal, you can send me the fit.list object and I can
   try to debug.

   Hi Francesco,
   Sorry I should have written:
   all.equal(target, reduce(target, min.gapwidth=0))
   ...as adjacent ranges are fine.
   It is probably easier to re-run counting after reducing the target with
   min.gapwidth=0.  There could be complicated overlapping ranges generated if
   you have overlapping but not identical regions in the original target file.


   1. mailto:love at molgen.mpg.de
   2. mailto:f.lescai at ucl.ac.uk

More information about the Bioconductor mailing list