Linear mixed model SNP h2

[kpric...@gmail.com]

Hi,

I have a small sib-pair sample I did a GWAS on. I used linear mixed model and PC's. The fixed effects were phenotype~SNP + PCs and the random effects were the family relation. I read on the LDSC home page I cannot do SNP h2 on this sample. How would I calculate LD score regression? Should I only include one family member? Can I do partitioned h2 on the entire sample?

Number of SNPs: 1061901
Total Observed scale h2: -0.26 (0.76)
Lambda GC: 1.0225
Mean Chi^2: 1.0138
Intercept: 1.0171 (0.0074)
Ratio: 1.2387 (0.5346)
Z=<0

Thank you for your help,

Kaitlyn