Table 2.

Contributions of the “Genetics of Gene Expression and Phenotype” subgroup

			Real data
First author	Sample size	Relatedness correction	Phenotype	Genotype	Simulated data	Analytic approach	Results
Ainsworth [27]	R 638	GWAS: FaST-LMM	Covariate adjusted mean SBP, DBP	427,952 GWAS SNPs	Analyzed but not presented	Pairwise association, WGCNA to identify variables for causal modeling in SEM and BUF	Weak significance, high concordance between SEM and BUF
		WGCNA SEM, BUF: none
Pitsillides [24]	R 267	Linear mixed effects models	DBP, SBP	12,296,048 SNVs from WGS	Not used	Test of enrichment of cis-eQTL in known BP loci and regulatory regions; pairwise association of expression and BP	Many highly significant eQTL. Enrichment of eQTL in known BP loci and regulatory regions
Tong [26]	U 142	None needed	SBP, DBP, HTN, adjusted for covariates	6,956,910 SNVs from WGS in 17,558 genes	Not used	Similarity-based test for joint effects of genotype, gene expression, phenotype	Weak significance, but some benefit from using genotypes and gene expression
Radkowski [25]	R 340	None made	HTN at several time points; change in BP	Not used	Not used	Change in BP (in individuals with no HTN) modeled as a function of gene expression and covariates	7 potentially predictive HTN gene expression probes identified in 6 genes

BP blood pressure; BUF Bayesian unified framework; DBP diastolic blood pressure; eQTL expression quantitative trait locus; GWAS genome-wide association study; HTN hypertension; R related individuals; SBP systolic blood pressure; SEM structural equation modeling; SNP single nucleotide polymorphism; SNV single nucleotide variant; U unrelated individuals; WGCNA weighted gene correlation network analysis; WGS whole-genome sequencing