Compute TWAS weights via ridge regression (glmnet, alpha = 0)

Convenience wrapper for glmnetWeights with alpha = 0.

Usage

ridgeWeights(
  X,
  y,
  fitRetention = c("none", "slim", "full"),
  methodArgs = GlmnetOptions()
)

Arguments

X

Numeric genotype / design matrix (samples x variants).

y

Numeric response (phenotype) vector of length nrow(X).

fitRetention

How much of the fit is kept as the "fit" attribute of the returned weights: "none" (default) attaches nothing; "slim" the selected fit, as a list of its coefficients (coef, over the variants fitted), intercept, lambda and cross-validated error there (cvm, cvsd); "full" the whole cv.glmnet object, with every lambda's coefficients. Either names the variants it was fitted on.

methodArgs

Additional arguments forwarded to glmnet::cv.glmnet, built with GlmnetOptions. A bare list is refused, since it cannot be checked against the engine's formals.

Value

Numeric vector of length ncol(X) of per-variant weights, with the fit attached as attribute "fit" unless fitRetention = "none".

Examples

data(eqtlRegionExample)
X <- eqtlRegionExample$X[, 1:30]
y <- eqtlRegionExample$yRes
ridgeWeights(X, y)
#>                [,1]
#>  [1,] -5.062171e-04
#>  [2,] -5.062365e-04
#>  [3,]  3.735322e-04
#>  [4,] -3.593713e-05
#>  [5,] -3.593628e-05
#>  [6,] -1.674700e-04
#>  [7,] -2.100183e-04
#>  [8,]  1.837077e-04
#>  [9,] -1.703815e-04
#> [10,] -1.704199e-04
#> [11,] -5.292013e-05
#> [12,] -2.994142e-04
#> [13,] -5.664245e-05
#> [14,] -1.482979e-04
#> [15,] -5.291703e-05
#> [16,]  1.837147e-04
#> [17,] -3.250155e-04
#> [18,] -3.581537e-04
#> [19,]  6.945953e-05
#> [20,] -3.581769e-04
#> [21,] -3.581882e-04
#> [22,] -1.674813e-04
#> [23,] -3.581882e-04
#> [24,]  1.725879e-05
#> [25,] -5.667378e-05
#> [26,] -1.674818e-04
#> [27,] -3.511836e-04
#> [28,] -7.401435e-05
#> [29,]  5.302680e-05
#> [30,]  2.284702e-04