Open xEmz opened 4 months ago
@xEmz Computing LD scores takes 1-2 hours per chromosome, can can be slightly sped up by using multiple cores. You should use a compute cluster to parallelize across chromosomes.
One other thing I notice is that you should use --ld-wind-kb 1000
to match the typical way LD scores are computed.
Hello, I want to analyse African GWAS data with LDSC. However, I have been unable to find LD scores with African ancestry online. I have generated LD scores using 1000G data, but I find heritability analysis to take very long with this.
Does anyone know about African LD scores or how to run it more efficiently?
Settings used to make LD scores:
python ldsc.py --bfile /references/african_1k3v5_plink/1000Gp3v5.20130502.AFR.chr1 --l2 --ld-wind-kb 1 --out /references/afr_w_ld_chr/1