Subject: How to compute Q_SNP for a higher-order factor model in GenomicSEM?

[Goodluck li]

Hello GenomicSEM team,

My current model is:

F1 =~ 1*T1 + T2 + T3 + T4 + T5 F2 =~ 1*T6 + T7 + T8 G =~ 1*F1 + F2 G ~ SNP

I’m running:

result <- userGWAS( model = model_CP, SNPs = test_SNPs, estimation = "DWLS", sub = c("G ~ SNP") )

I noticed that Q_SNP cannot be computed in this setup (the Q/Q_SNP column returns NA).
In this situation, is it acceptable to proceed without computing Q_SNP—i.e., to not filter GWAS hits by Q_SNP? If Q_SNP-based filtering is required, how should I compute it for a higher-order (second-order) factor model? If possible, could you share example code?

Many thanks!

Best regards,

goodluck