Hello,
I ran my analysis on GISTIC_2.0 version 6.15.30 on Gene Pattern.
I have 58 samples of whole genome sequencing data with the seg copy number file. I prepared the input for gistic as stated in the literature and the forum: a txt file with 6 columns: sample, chromosome, start, end, marker and seg.CN. I left the marker column with NA, that is how I read it here in the forum and the Seg.CN was calculated by using log2(CN) - 1.
It looked like this:
head(input_gistic)
sample chromosome start end marker Seg.CN
2010397T X 1 42514345 NA -0.981793143
2010397T 21 38034611 38039193 NA -0.003683571
2010397T 12 76569304 76569305 NA 0.962771101
2010397T X 47540672 47545964 NA -1.074304339
2010397T 12 104323150 104323189 NA -0.824826734
2010397T X 42514346 47540671 NA 0.021906077
Then I run gistic on Gene pattern with the following values:
run.analysis(gp,
"urn:lsid:broad.mit.edu:cancer.software.genepattern.module.analysis:00125:6.15.30",
refgene.file="Human_Hg19.mat",
seg.file="peritoneum.txt",
markers.file="",
maxspace="",
array.list.file="",
cnv.file="",
gene.gistic="1",
amplifications.threshold="0.5",
deletions.threshold="-0.5",
join.segment.size="",
q-value.threshold="",
remove.X="1",
cap.val="2.5",
confidence.level="0.90",
run.broad.analysis="1",
focal.length.cutoff="0.50",
max.sample.segs="4000",
arm.peel="1",
sample.center="",
gene.collapse.method="extreme",
output.prefix="peritoneum",
logdat="")
Everything went fine and I got the results.
But I am not sure if the results are right and if the method I am using to check the results is right.
So for example I have an Amplification peak in 22q12.1 with the wide peak limits at chr22:29065860-29065888(probes 325738:325749) in 18 samples. But If I go back to the segmented copy number file and look at this samples and this chromosome I can not find the amplification at those parts of the chromosome.
So now I am not sure if that is generally not possible because the algorithm is based on something else or if something in my input or the analysis went wrong.
Input would be really appreciated!!!!
EDIT: Never mind, you already posted started a thread there 2 days ago, so the problem is that you're opening multiple unconnected threads on different groups for the same question because (I'm assuming here) you want a quick response. On a weekend. Do you think such an inconsiderate action will come across well to people that are on both communities?
I think GenePattern-GISTIC has its own Google group, you might be better off asking there.