cBioPortal survival analysis help

Question

[doc.golly]

dear cBioPortal team and users,

In cBioPortal's survival analysis (web portal), when stratifying samples into "with/without Cases with Alteration(s) in Query Gene(s)" based on a gene list, how is the overall with/without Cases with Alteration(s) groups in Query Gene(s) stratified (/calculated?

For example in http://bit.ly/2a4BeLT Select Genomic Profiles: mRNA Expression z-Scores (RNA Seq V2 RSEM): Enter a z-score threshold ±:2 Tumor Samples with mRNA data (RNA Seq V2) (498 samples) / 40 Genes

I assume alteration here is mRNA Upregulation vs mRNA Downregulation for each gene. So, for each gene each sample has a z-score (indicate relative over- or under-expression for each gene in comparison to the rest of the cohort), but I am not sure where to go from here when it comes to dividing the samples into two distinct groups.

all the best, doc golly

Answer 1

[edco84]

Alteration means that the Z score of that gene is 2 standard deviations away from the Z scores of the selected data. You can try playing around this cutoff.