Hi,

I am trying to replicate the heritability estimates based on the insomnia GWAS summary statistics using LDSC. However, I have encountered a problem as my estimates seem to be only about half of the original estimates listed in Table S1. Despite my efforts to locate the error, I have been unsuccessful in finding the source of the problem.

The following table list the original UKB estimates and my own estimates based on their public summary statistics.

Sex h2 (Table S1) h2_SE (Table S1) h2 (my calculation) h2_SE (my calculation)

Both (UKB) 0.0823 0.0036 0.0458 0.0021

Male (UKB) 0.0887 0.007 0.0438 0.0035

Female (UKB) 0.0906 0.0051 0.0542 0.0032

Below are the commands I used for female estimates.

zcat insomnia_female_ukb2b_EUR_sumstats_20190311_with_chrX_mac_100.txt.gz |cut -f 6,11,12,15,16,17 -d $'\t'|sed  "1s/STAT/Z/" > female_useful.txt
../munge_sumstats.py --sumstats  female_useful.txt --N-col NMISS --snp RSID_UKB   --a1 A1_UKB  --a2 A2_UKB --out female_insomnia_GWAS_2022 --merge-alleles LDSCORE_w_hm3.snplist --chunksize 500000
../ldsc.py --h2 female_insomnia_GWAS_2022.sumstats.gz --ref-ld-chr eur_w_ld_chr/ --w-ld-chr eur_w_ld_chr/ --out merged_GWAS_results_h2

The GWAS summary statistics for the female heritability were downloaded from

https://ctg.cncr.nl/documents/p1651/insomnia_female_ukb2b_EUR_sumstats_20190311_with_chrX_mac_100.txt.gz

I have tried to contact the first and corresponding authors but they seems to be too busy to reply. Can anyone tell me where am I wrong? Thanks.