Delta z score between control and treatment

Question

[bioinfonerd]

Hi everyone,

I have RNA expression data for 20 control and 20 matched treatment samples. I need to plot z-score change for a gene signature between control and treatment for all the 20.

This is what I think to do:

1. Log2 of the expression values
2. Calculate z-scores separately for all control
3. In the same way calculate z scores for all treatment samples
4. For a gene subtract z-score of treatment from z-score of control to get the change
5. Plot the z-score change per sample for a gene

Is it ok to calculate z-score separately for control send treatment or should I calculate the z-score together for them? Please let me know the correct way to handle this.

Thanks!

Answer 1

[Prakash]

Z score is generally used to normalize value to mean and to scale data in some range (depends upon standard deviation) . The best way to compare samples (control and treatment) is to calculate across all the sample.

[bioinfonerd]

Thanks very much! That's helpful.