P-values of GWAS summary statistics

[koh2ng0]

Hello,

I have downloaded the GWAS summary statistics for the 74 traits and diseases used in your analysis from Figshare. Upon reviewing the files to rerun MAGMA, I noticed that while CHISQ and Z values are present, there are no p-values for each SNP. As far as I know, p-values are necessary for accurate MAGMA calculations. Could you clarify which values you used when running MAGMA? If possible, would it be feasible to provide GWAS summary statistics that include p-values?

Thank you in advance for your assistance!

[martinjzhang]

Hi, the p-values can be computed from chi2 statistics using a chi-square test with 1 degree of freedom

[koh2ng0]

Thanks for your nice reply! I will try

Best regards,
Koh

[koh2ng0]

Hi, the p-values can be computed from chi2 statistics using a chi-square test with 1 degree of freedom

I reviewed your reply and have been calculating p-values from CHISQ values. However, I noticed that for several GWAS summary statistics, CHISQ values are not provided, and only Z-scores are available (e.g., PASS_BIP_Mullins2021.sumstats). How do you think I should handle these cases?

Thank you.

[martinjzhang]

Hi, CHISQ = Z^2