Exception Error when computing non-paired CBS: Exception: .. Too many non-finite values ......

[sean nj]

Hi, it's me again :)

 

I have some level 2 TCGA agilent CGH 244A data, the text files of log2 ratios. I followed vignettes (http://aroma-project.org/node/43 and http://www.aroma-project.org/node/88) and created UGP files and data files (asb files) containing log2 CN ratios. But I got the error when I ran CBS computing by  > fit(sm, verbose=-10). Basically it complianed that there are too many non-finite values on chromosome 1 of assay sample 1:

      Array #1 ('MSK_00001_251469322729') of 6 on chromosome 1...
      Error in getRawCnData.CopyNumberChromosomalModel(this, ceList = ceList,  :
 
      [2012-05-18 14:02:13] Exception: Something is wrong with the copy-number ratios of sample 'MSK_00001_251469322729,S01_CGH-v4_91__GCN_V3,log2ratio,total' relative to reference '      .average-signals-median-mad,78c9a8b043bc37fd29db00f3f674295c' on chromosome 1. Too many non-finite values: 6721 (34.9% > 20.0%) out of 19246.

I also have one more question regarding CBS. Once I finish CBS computing, I end up with a text file such as:

 

         sample chromosome start stop mean count
      1 NA06991 1 61736 106013377 -0.002 65870
      2 NA06991 1 106019206 106022376 -1.675 1
      3 NA06991 1 106024056 149036525 -0.002 11462
      4 NA06991 1 149040066 149256692 -0.443 141
      5 NA06991 1 149259417 149436843 -0.144 36

 

Is there a way to map the original probe ID or gene IDs to segmentation result? I want to end up with a table that has a column for original probe ID, a column for the related CN ratio mean from CBS, a column for chromosome, and a column for chromosome start position.

 

Thanks a lot for the help!

 

Ying

 

 

> library(aroma.core)

> library(aroma.cn)

> chipType <- "HG-CGH-244A,TCGA_GBM_MSK"
> unf <- TextUnitNamesFile$byChipType(chipType)
> ugp <- getAromaUgpFile(unf)

> dataSet <- "TCGA_GBM_MSKCC"
> ds <- AromaUnitTotalCnBinarySet$byName(dataSet, chipType=chipType)
> print(ds)
AromaUnitTotalCnBinarySet:
Name: TCGA_GBM_MSKCC
Tags:
Full name: TCGA_GBM_MSKCC
Number of files: 6
Names: MSK_00001_251469322729, MSK_00003_251469322731, MSK_00006_251469322732, ..., MSK_00010_251469322627 [6]
Path (to the first file): rawCnData/TCGA_GBM_MSKCC/HG-CGH-244A,TCGA_GBM_MSK
Total file size: 5.40 MB
RAM: 0.01MB
> sm <- CbsModel(ds)
Loading required package: DNAcopy
> print(sm)
CbsModel:
Name: TCGA_GBM_MSKCC
Tags:
Chip type (virtual): HG-CGH-244A
Path: cbsData/TCGA_GBM_MSKCC/HG-CGH-244A
Number of chip types: 1
Sample & reference file pairs:
Chip type #1 ('HG-CGH-244A') of 1:
Sample data set:
AromaUnitTotalCnBinarySet:
Name: TCGA_GBM_MSKCC
Tags:
Full name: TCGA_GBM_MSKCC
Number of files: 6
Names: MSK_00001_251469322729, MSK_00003_251469322731, MSK_00006_251469322732, ..., MSK_00010_251469322627 [6]
Path (to the first file): rawCnData/TCGA_GBM_MSKCC/HG-CGH-244A,TCGA_GBM_MSK
Total file size: 5.40 MB
RAM: 0.01MB
Reference data set/file:
<average across arrays>
RAM: 0.00MB
> fit(sm, verbose=-10)

Building tuple of reference sets...
 No reference available.
 Calculating average copy-number signals...
  Retrieving average unit signals across 6 arrays...
   Searching for an existing file...
    Paths:
    [1] "rawCnData/TCGA_GBM_MSKCC/HG-CGH-244A,TCGA_GBM_MSK"
    Filename: .average-signals-median-mad,78c9a8b043bc37fd29db00f3f674295c.asb
    Searching path #1 of 1...
     Path: rawCnData/TCGA_GBM_MSKCC/HG-CGH-244A,TCGA_GBM_MSK
     Pathname: rawCnData/TCGA_GBM_MSKCC/HG-CGH-244A,TCGA_GBM_MSK/.average-signals-median-mad,78c9a8b043bc37fd29db00f3f674295c.asb
     Found an existing file.
    Searching path #1 of 1...done
    Located pathname: rawCnData/TCGA_GBM_MSKCC/HG-CGH-244A,TCGA_GBM_MSK/.average-signals-median-mad,78c9a8b043bc37fd29db00f3f674295c.asb
   Searching for an existing file...done
   Loading existing data file...
   Loading existing data file...done
   AromaUnitTotalCnBinaryFile:
   Name: .average-signals-median-mad
   Tags: 78c9a8b043bc37fd29db00f3f674295c
   Full name: .average-signals-median-mad,78c9a8b043bc37fd29db00f3f674295c
   Pathname: rawCnData/TCGA_GBM_MSKCC/HG-CGH-244A,TCGA_GBM_MSK/.average-signals-median-mad,78c9a8b043bc37fd29db00f3f674295c.asb
   File size: 921.57 kB (943689 bytes)
   RAM: 0.00 MB
   Number of data rows: 235834
   File format: v1
   Dimensions: 235834x1
   Column classes: double
   Number of bytes per column: 4
   Footer: <createdOn>20120518 13:04:05 EDT</createdOn><platform>Agilent</platform><chipType>HG-CGH-244A</chipType><srcDetails><nbrOfFiles>6</nbrOfFiles><checkSum>5629a7bc8e75693ad556a0d8c9f3417c</checkSum></srcDetails><params><meanName>median</meanName><sdName>mad</sdName></params>
   Platform: Agilent
   Chip type: HG-CGH-244A
   Number of units to be updated: 7
   Processing chunk...
     chr "Indices in chunk:"
     int [1:7] 93672 120223 160149 216186 234895 234902 234936
    Reading data...
    Reading data...done
    Estimating averages and standard deviations...
    Estimating averages and standard deviations...done
    Writing estimates...
    Writing estimates...done
   Processing chunk...done
  Retrieving average unit signals across 6 arrays...done
 Calculating average copy-number signals...done
Building tuple of reference sets...done
Using reference tuple:
AromaUnitTotalCnBinarySetTuple:
Name: TCGA_GBM_MSKCC
Tags:
Chip types: HG-CGH-244A
AromaUnitTotalCnBinarySet:
Name: TCGA_GBM_MSKCC
Tags:
Full name: TCGA_GBM_MSKCC
Number of files: 6
Names: .average-signals-median-mad, .average-signals-median-mad, .average-signals-median-mad, ..., .average-signals-median-mad [6]
Path (to the first file): rawCnData/TCGA_GBM_MSKCC/HG-CGH-244A,TCGA_GBM_MSK
Total file size: 5.40 MB
RAM: 0.01MB
RAM: 0.00MB
Extract DataFileMatrix...
 Array: 1
 Test data sets:
 AromaUnitTotalCnBinarySetTuple:
 Name: TCGA_GBM_MSKCC
 Tags:
 Chip types: HG-CGH-244A
 AromaUnitTotalCnBinarySet:
 Name: TCGA_GBM_MSKCC
 Tags:
 Full name: TCGA_GBM_MSKCC
 Number of files: 6
 Names: MSK_00001_251469322729, MSK_00003_251469322731, MSK_00006_251469322732, ..., MSK_00010_251469322627 [6]
 Path (to the first file): rawCnData/TCGA_GBM_MSKCC/HG-CGH-244A,TCGA_GBM_MSK
 Total file size: 5.40 MB
 RAM: 0.01MB
 RAM: 0.00MB
 Reference data sets:
 AromaUnitTotalCnBinarySetTuple:
 Name: TCGA_GBM_MSKCC
 Tags:
 Chip types: HG-CGH-244A
 AromaUnitTotalCnBinarySet:
 Name: TCGA_GBM_MSKCC
 Tags:
 Full name: TCGA_GBM_MSKCC
 Number of files: 6
 Names: .average-signals-median-mad, .average-signals-median-mad, .average-signals-median-mad, ..., .average-signals-median-mad [6]
 Path (to the first file): rawCnData/TCGA_GBM_MSKCC/HG-CGH-244A,TCGA_GBM_MSK
 Total file size: 5.40 MB
 RAM: 0.01MB
 RAM: 0.00MB
 Test data files:
 $`HG-CGH-244A`
 AromaUnitTotalCnBinaryFile:
 Name: MSK_00001_251469322729
 Tags: S01_CGH-v4_91__GCN_V3,log2ratio,total
 Full name: MSK_00001_251469322729,S01_CGH-v4_91__GCN_V3,log2ratio,total
 Pathname: rawCnData/TCGA_GBM_MSKCC/HG-CGH-244A,TCGA_GBM_MSK/MSK_00001_251469322729,S01_CGH-v4_91__GCN_V3,log2ratio,total.asb
 File size: 921.40 kB (943513 bytes)
 RAM: 0.00 MB
 Number of data rows: 235834
 File format: v1
 Dimensions: 235834x1
 Column classes: double
 Number of bytes per column: 4
 Footer: <createdOn>20120518 11:04:29 EDT</createdOn><platform>Agilent</platform><chipType>HG-CGH-244A</chipType>
 Platform: Agilent
 Chip type: HG-CGH-244A
 
 attr(,"class")
 [1] "AromaUnitTotalCnBinaryFileList" "GenericDataFileList"            "list"                         
Extract DataFileMatrix...done
Genomic-signal tags: S01_CGH-v4_91__GCN_V3,log2ratio,total
Reference tags: 216b0999d2e57831383f208942c5d3c6
Array #1 ('MSK_00001_251469322729') of 6 on chromosome 1...
Error in getRawCnData.CopyNumberChromosomalModel(this, ceList = ceList,  :
 
[2012-05-18 14:02:13] Exception: Something is wrong with the copy-number ratios of sample 'MSK_00001_251469322729,S01_CGH-v4_91__GCN_V3,log2ratio,total' relative to reference '.average-signals-median-mad,78c9a8b043bc37fd29db00f3f674295c' on chromosome 1. Too many non-finite values: 6721 (34.9% > 20.0%) out of 19246.

  at #06. getRawCnData.CopyNumberChromosomalModel(this, ceList = ceList,
              refList = rfList, chromosome = chromosome, ..., estimateSd = FALSE,
              verbose = less(verbose))
          - getRawCnData.CopyNumberChromosomalModel() is in environment 'aroma.core'

  at #05. getRawCnData(this, ceList = ceList, refList = rfList, chromosome = chromosome,
              ..., estimateSd = FALSE, verbose = less(verbose))
          - getRawCnData() is in environment 'aroma.core'

  at #04. extractRawCopyNumbers.CopyNumberChromosomalModel(this, array = array,
              chromosome = chr, maxNAFraction = maxNAFraction)
          - extractRawCopyNumbe
In addition: Warning messages:
1: In log(M, base = 2) : NaNs produced
2: In log(A, base = 2) : NaNs produced
Array #1 ('MSK_00001_251469322729') of 6 on chromosome 1...done
>
> sessionInfo()
R version 2.15.0 (2012-03-30)
Platform: x86_64-pc-mingw32/x64 (64-bit)

locale:
[1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252    LC_MONETARY=English_United States.1252
[4] LC_NUMERIC=C                           LC_TIME=English_United States.1252   

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base    

other attached packages:
 [1] aroma.cn_1.0.0         DNAcopy_1.30.0         aroma.affymetrix_2.5.0 affxparser_1.28.0      aroma.apd_0.2.2      
 [6] R.huge_0.3.0           aroma.core_2.5.0       aroma.light_1.24.0     matrixStats_0.5.0      R.rsp_0.7.5          
[11] R.cache_0.6.2          R.filesets_1.1.5       digest_0.5.2           R.utils_1.12.1         R.oo_1.9.3           
[16] R.methodsS3_1.2.2    
>

[Henrik Bengtsson]

Hi.

On Fri, May 18, 2012 at 11:17 AM, sean nj <njs...@gmail.com> wrote:
> Hi, it's me again :)
>
> I have some level 2 TCGA agilent CGH 244A data, the text files of log2
> ratios. I followed vignettes (http://aroma-project.org/node/43 and
> http://www.aroma-project.org/node/88) and created UGP files and data files
> (asb files) containing log2 CN ratios. But I got the error when I ran CBS
> computing by  > fit(sm, verbose=-10). Basically it complianed that there are
> too many non-finite values on chromosome 1 of assay sample 1:
>       Array #1 ('MSK_00001_251469322729') of 6 on chromosome 1...
>       Error in getRawCnData.CopyNumberChromosomalModel(this, ceList =
> ceList,  :
>
>       [2012-05-18 14:02:13] Exception: Something is wrong with the
> copy-number ratios of sample
> 'MSK_00001_251469322729,S01_CGH-v4_91__GCN_V3,log2ratio,total' relative to
> reference '
> .average-signals-median-mad,78c9a8b043bc37fd29db00f3f674295c' on chromosome
> 1. Too many non-finite values: 6721 (34.9% > 20.0%) out of 19246.

As I understand it, you're imported CN signals are already ratios,
i.e. ratios relative to a reference. When you do:

sm <- CbsModel(ds)

the CbsModel code is not clever enough to know this, so it will do use
the default and pool across all samples in 'ds' and use their average
as a reference, i.e. divide your signals with. To avoid this, there
is an (yet to become official) setting you can use as follows:

sm <- CbsModel(ds)
sm$.calculateRatios <- FALSE

See if that helps.

/Henrik

[snip]

> --
> When reporting problems on aroma.affymetrix, make sure 1) to run the latest
> version of the package, 2) to report the output of sessionInfo() and
> traceback(), and 3) to post a complete code example.
>
>
> You received this message because you are subscribed to the Google Groups
> "aroma.affymetrix" group with website http://www.aroma-project.org/.
> To post to this group, send email to aroma-af...@googlegroups.com
> To unsubscribe and other options, go to http://www.aroma-project.org/forum/