How to perform GISTIC analysis on GenePattern?

Question

[Dmitri Ivanovich]

Hi guys, I have got a CNV data from TCGA and the marker file as shown below,

• Seg file:

• Sample Chromosome Start End Num_Probes Segment_Mean

• TCGA.05.4249.01A 1 3218610 120527361 67456 -0.1725

• TCGA.05.4249.01A 1 149881398 167526508 10663 0.5859
• TCGA.05.4249.01A 1 167526675 167526823 2 -1.1518
• TCGA.05.4249.01A 1 167526972 247813706 50571 0.5816

• TCGA.05.4249.01A 2 484222 242476062 130861 0.0585

• Markers file:

• Probe.Name Chromosome Start

• CN_473963 1 61735
• CN_473964 1 61808
• CN_473965 1 61823
• CN_477984 1 62152
• CN_473981 1 62920
• CN_473982 1 62937

And my goal is to perform GISTIC analysis with the GISTIC 2.0 module in GenePattern, but the result is always like this: "GISTIC version 2.0.23 GISTIC 2.0 input error detected: 76606 segment start or end positions in '/opt/gpcloud/gp_home/users/genye/uploads/tmp/run8835511072592266907.tmp/seg.file/1/biguoshu1.txt' do not match any markers in '/opt/gpcloud/gp_home/users/genye/uploads/tmp/run6032895478607754030.tmp/markers.file/2/markersMatrix.txt'. First bad position is 10:24732567 at line 33."

I have uploaded my files in .txt format and choose the GISTIC version 6.15.28 and Human_hg19.mat as the refgene file. All other parameters were by default. Could anyone please tell me what the problem is and how to solve? Thank you !

[2953531955]

Hi,did you fix the problem? I had the same problem

[Kevin Blighe]

Please show the exact error message, a sample of your input data, and all commands that you have tried. Thank you.

[2953531955]

thank you for your reply!

error: GISTIC 2.0 input error detected:
198278 segment start or end positions in '/opt/gpcloud/gp_home/users/yuduoduo/uploads/tmp/run1102255665308516451.tmp/seg.file/1/MaskedCopyNumberSegment.txt' do not match any markers in '/opt/gpcloud/gp_home/users/yuduoduo/uploads/tmp/run5183191344025581000.tmp/markers.file/2/genome.info.6.0_hg19.na31_minus_frequent_nan_probes_sorted_2.1(1).txt'.
First bad position is 1:2116145 at line 1.

input file:

TCGA-MQ-A4LJ-01A    1   62920   2116145 358 0.0051
TCGA-MQ-A4LJ-01A    1   2125269 3259074 359 -0.0884
TCGA-MQ-A4LJ-01A    1   3259896 12779433    5732    0.0459
TCGA-MQ-A4LJ-01A    1   12792599    12922922    33  -0.7534

marker file:

genome.info.6.0_hg19.na31_minus_frequent_nan_probes_sorted_2.1.txt
 CN_473963  1   61735
CN_473964   1   61808
CN_473965   1   61823
CN_477984   1   62152
CN_473981   1   62920
CN_473982   1   62937
CN_497980   1   72704

All other parameters were by default thank you !

[Kevin Blighe]

You could try without the markers file, which is now possible with later versions of GISTIC. Also, just double-check that the formatting of your files is correct.

[2953531955]

Thank you for you help Kevin! but another problem arised:I got many regions amplified/deleted. The plots are very noisy, amplification/deletion occurred in almost every gene.Could you please tell me what the problem is and how to solve? Thank you !

[Kevin Blighe]

Hi, you are not giving me much information with which I could use to begin to help. Please share, in detail, the data that you obtained, and the code that you used to process it.

[2953531955]

sorry! The plot is always like this: https://ibb.co/swrq1qD

my maskedCNVsegment data was from TCGA and I perform GISTIC analysis with the GISTIC 2.0 module in GenePattern. All other parameters were by default .The plots are very noisy, amplification/deletion occurred in almost every gene

[Kevin Blighe]

You took data from the GDC? That data is segmented copy number data produced by DNAcopy, I believe. You then used that as input to GISTIC?

Could you take a look here to see how this matches up to what you have done? - A: How to extract the list of genes from TCGA CNV data

[Kevin Blighe]

Follow up on this part of the error:

First bad position is 10:24732567 at line 33.

As it alludes to chromosome 10, perhaps one of your files is not sorted numerically, and is instead sorted lexicographically

[Dmitri Ivanovich]

Thank you for you helping Kevin! But after I sorted my files numerically, it still showed similar result (... do not match any markers...) . Is it possible that the Marker File I submitted doesn't fit TCGA data, or that the online version of GISTIC2 doesn't work at all?

[Kevin Blighe]

Perhaps you can contact the GISTIC team directly:

• A: GISTIC input parameters.
• A: GISTIC 2.0 all threshold file versus genome browser

[Kevin Blighe]

I think that I read somewhere, by the way, that the IDs have to be like this:

TCGA-MQ-A4LJ

So, less the final part. Can you try?