In the provided RNASeq dataset, there is a zScore column. In the Data Walkthrough page for this data, this column is described as:
"normalized counts (counts for a gene can vary significantly due to technical differences in different studies, so we normalized the counts into a zScore for a comparable measurement between samples for the same gene)"
Do you have any specifics on how this zScore was calculated?
Thank you.
Created by Jeff Green jeff_green Hi @lars.ericson ,
Sorry for not being explicit. When I say 'per-sample' I meant that we take the mean of all the genes *in* a sample, so each Z-score represents the relative expression of that gene in a sample. A positive score indicates a gene expression was higher than the mean for that sample, and a negative value means it's lower than the mean of that sample. This was done for all samples, regardless of sample origin.
-sara @sgosline do you take the mean of all the samples e.g. all tumor types and normal? What does a negative score mean? The Z-score is calculated on a per-sample basis, where we subtract the mean of the sample counts from the samples counts and divide by the standard deviation. For more information you can review it [here](https://en.wikipedia.org/wiki/Standard_score).
-sara cc @sgosline