stacked_prs: Stacked P+T polygenic risk scores by regularized regression

In vincent10kd/polygenic: Efficient and User-friendly P+T Polygenic Risk Score Creation with Summary Statistics

Stacked P+T polygenic risk scores by regularized regression

Description

Uses individual-level data to combine ('stack') polygenic risk scores based on different tuning parameter combinations to optimize predictive performance. Based on Privé et al. AJHG 2019.

Usage

stacked_prs(
  prs_grid,
  y,
  nfolds = 10,
  family = "binomial",
  plot_error = TRUE,
  seed = NULL,
  ...
)

Arguments

prs_grid	The element 'result_df' from the list output by prs_grid().
y	The observed outcome (trait).
nfolds	Number of folds to use in k-fold cross-validation within cv.glmnet().
family	The distribution family of the outcome, e.g. binomial, gaussian.
plot_error	If TRUE, plots the error of the stacked PRS by the tuning parameter lambda.
seed	If not NULL, sets a seed prior to running cv.glmnet().
...	Additional arguments to pass to cv.glmnet().

Value

A list containing the model, the stacked PRS, and the non-zero model coefficients.

vincent10kd/polygenic documentation built on Feb. 25, 2024, 10:17 a.m.