Re: Qestions for "The cBio Cancer Genomics Portal: An Open Platform for Exploring Multidimensional Cancer Genomics Data"

[Nikolaus Schultz]

Dear Yanyan,

Here is an explanation on how the thresholds are set:

The table in all_thresholded.by_genes.txt (which is the part of the GISTIC output that is used to determine the copy-number status of each gene in each sample in cBioPortal) is obtained by applying both low- and high-level thresholds to to the gene copy levels of all the samples. The entries with value +/- 2 exceed the high-level thresholds for amps/dels, and those with +/- 1 exceed the low-level thresholds but not the high-level thresholds. The low-level thresholds are just the 'amp_thresh' and 'del_thresh' noise threshold input values to GISTIC (typically 0.1 or 0.3) and are the same for every thresholds.

By contrast, the high-level thresholds are calculated on a sample-by-sample basis and are based on the maximum (or minimum) median arm-level amplification (or deletion) copy number found in the sample. The idea, for deletions anyway, is that this level is a good approximation for hemizygous given the purity and ploidy of the sample. The actual cutoffs used for each sample can be found in a table in the output file sample_cutoffs.txt.  All GISTIC output files for TCGA are available at:

gdac.broadinstitute.org

If you want to learn more about GISTIC please refer to :

Mermel C, Schumacher S, et al. (2011). "GISTIC2.0 facilitates sensitive and confident localization of the targets of focal somatic copy-number alteration in human cancers." Genome Biology,12:R41.

If you have any further questions, you can write to me directly.

Regards,

Andrew 

___________________________
Andrew Cherniack, PhD                                                
Research Scientist II
Broad Institute
7 Cambridge Center
Cambridge, Mass 02142
email: ache...@broadinstitute.org
___________________________

On May 29, 2013, at 10:02 AM, 平艳艳 <pingy...@yahoo.com.cn> wrote:

Dear Sirs and Madams,

 I am a student majoring in bioinformatics in Harbin Medical University of China. Recently, I have read your paper "The cBio Cancer Genomics Portal: An Open Platform for Exploring Multidimensional Cancer Genomics Data", and I went to http://www.cbioportal.org/public-portal/index.do to view it. It is a useful tool for me. I have some questions: 1. The web provides the results about the "Putative copy-number alterations from GISTIC", I want to know which data these results is derived from, the aCGH or SNP 6.0? 2. I know that the GISTIC can calculated the significant copy nubmer alteration across samples. But how to determined the altered status of genes in each samples, -2,-1,0,1,2, what is the threshold? Whether these thresholds were applied to the means values of segments? I have carefully read the FAQ, but I still don't find the answer.

 

I am looking forward for your reply.

 

Best wishes.

 

Sincerly Yours,

 

Yanyan Ping

[David Tamborero]

I do not know whether this is the best site to ask for this question, but it worths the shot;  I am not able to elucidate whether a '0' from the Gistic copy number resulting state means indeed that the gene/sample is diploid or it means that no copy number change exists as compared with the normal pair.

In other words, if the sample has a 'germline loss of an allele', the Gistic takes into account this issue somehow?

 Thanks in advance!

[Nikolaus Schultz]

Hi David,

If the copy-number is from Affymetrix SNP6, which is the case for most of the TCGA samples, it is relative to a set of reference samples. The matched normal is not taken into account.

Niki.

--
You received this message because you are subscribed to the Google Groups "cBioPortal for Cancer Genomics Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cbioportal+...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[David Tamborero]

got it, thanks!