how to calculate variance in summary data.frame ?

[czlcaltech]

Hello, iDEA team:

Thanks for your iDEA packages. I read your iDEA paper published in NC, and I want to apply your package to analysis my data. However, when following the tutorial, I don't quite understand what is the variance (which is the second column in input summary data) refers to and how to calculate it, since the findmarkers function in Seurat and the data.frame returned from MAST do not include "variance". And due to my poor statistical knowledge, I cannot understand your explanations in the original paper. Could you please specify how the variance is calculated? Thanks.

Best,
Zhangliang

[YingMa0107]

Hi Zhangliang (@czlcaltech),

Sorry about the late reply. Here is an example to calculate the variance

# Assume you have obtained the MAST's results from Seurat, res_mast, which contains the column Pr(>Chisq) or LogFC. 
pvalue <- res_mast$`Pr(>Chisq)` #### the pvalue column
zscore <- qnorm(pvalue/2.0, lower.tail=FALSE) #### convert the pvalue to z-score
beta <- res_mast$LogFC ## effect size
se_beta <- beta/zscore ## to approximate the standard error of beta
var_beta = se_beta^2  ### square 

Hope this helps!

Best,
Ying