Generalized Least Squares

Runs GLS regression on genetic parameters. Drop-in replacement for GenomicSEM::summaryGLS(): returns a numeric matrix with columns betas, pvals, SE, Z and row names b0, b1, ... (intercept first when INTERCEPT = TRUE), and prints it before returning.

Usage

summaryGLS(OBJECT = NULL, Y = NULL, V_Y = NULL, PREDICTORS, INTERCEPT = TRUE)

Arguments

OBJECT

LDSC result object (used to extract Y and V_Y if provided)

Y

Response vector (vech of genetic covariance matrix)

V_Y

Covariance matrix of Y

PREDICTORS

Predictor matrix (n x p)

INTERCEPT

Include intercept (default TRUE)

Value

A numeric matrix with columns betas, pvals, SE, Z, matching GenomicSEM::summaryGLS.

Examples

# GLS regression of a synthetic response on two predictors.
Y <- c(0.50, 0.30, 0.20, 0.10)
V_Y <- diag(4) * 0.01
X <- matrix(c(1, 2, 3, 4,
              0, 1, 0, 1), nrow = 4, ncol = 2)
fit <- summaryGLS(Y = Y, V_Y = V_Y, PREDICTORS = X, INTERCEPT = TRUE)
#>     betas        pvals        SE          Z
#> b0  0.600 9.633570e-07 0.1224745  4.8989795
#> b1 -0.125 1.241933e-02 0.0500000 -2.5000000
#> b2 -0.025 8.230633e-01 0.1118034 -0.2236068
fit   # matrix with rows b0, b1, b2 and columns betas, pvals, SE, Z
#>     betas        pvals        SE          Z
#> b0  0.600 9.633570e-07 0.1224745  4.8989795
#> b1 -0.125 1.241933e-02 0.0500000 -2.5000000
#> b2 -0.025 8.230633e-01 0.1118034 -0.2236068