zhenin/HDL

Add support for comparison of genetic correlations with the block jackknife

Closed this issue · 2 comments

JoniColeman commented

Hi,

One of the benefits of LDSC is the option to output the jackknife delete values used to estimate rg, which can then be used to calculate robust differences between rg(A,B) and rg(C,D) (see e.g. https://groups.google.com/g/ldsc_users/c/uZyjiabAorc/m/fgX2SoEuBAAJ).

Would it be possible/meaningful to have some similar option in HDL?

Thanks,

Joni

zhenin commented

Hi Joni,

Yes, it could be a useful feature. Because s.e. in HDL is also from block-jackknife, so it can be easily added. I will add it in the next version of HDL.

Thanks for your interest in HDL!

Best regards,
Zheng

zhenin commented

Hello again,

A new optional argument jackknife.df has been added in HDL v1.3.8. You can get the jackknife estimates by setting jackknife.df=TRUE in R or the command-line version. You can find more details here.

Thanks for your advice! I will close this issue. Please feel free to come back to me if there is any comment on HDL.

Best,
Zheng

  • 👍1