Entering edit mode
6 months ago
YL ▴ 10
I found that following formula asked by other colleagues but no answer:
zprime = 0.5 * log((1+rg) / (1-rg))
ci = zprime ± 1.96*se
ci = ((exp(2ci) - 1) / (exp(2ci) + 1))
(rg and se taken from the LDSC output)
I wonder if this is the correct way of calculating 95% CI from LDSC output?
Thank you very much!
Are you able to find an answer for this? I have the exact question!
Based on my personal thought, I think the se from LDSC output probably needs to be transferred using Delta method based on Fisher transformation, and then used the transformed se to do ci = zprime ± 1.96*se
Hopefully we will get confirmation soon