rolypoly_roll: Main rolypoly wrapper function.

In rolypoly: Identifying Trait-Relevant Functional Annotations

Description

The entry point for rolypoly analysis. If no expression data, we assume that we are running just the vegas score processing.

Usage

rolypoly_roll(rolypoly = NULL, gwas_data = NULL, block_annotation = NULL,
  block_data = NULL, ld_folder = NULL, bootstrap_iters = 50,
  outlier_threshold = -1, perform_cv = F, n_folds = 10,
  gwas_z_filter = -1, add_spline = F, n_knots = 1, add_poly = F,
  n_degree = 2, run_light = T, gwas_link_parallel = F,
  bootstrap_parallel = F, keep_model = F, keep_gwas = F, ...)

Arguments

rolypoly	Previous rolypoly run to parts of pipeline.
gwas_data	Gwas data for a trait, including snp annotations.
block_annotation	Start and end points for blocks
block_data	Information about blocks.
ld_folder	Folder with LD information.
bootstrap_iters	Number bootstrap iterations to perform for inference.
outlier_threshold	Set to positive if we want to run robusted regression.
perform_cv	If we want to interpret annotation effects do not set this to T. However, if our goal is prediction accuracy then set this to T.
n_folds	number of folds for cross validation
gwas_z_filter	Z-score filter for SNPs, helps prevent large effects biasing inference.
add_spline	If we want to fit a spline to maf.
n_knots	number of knots to add to the spline.
add_poly	If we want to fit a polynomial to maf.
n_degree	the degree of the polynomial.
run_light	if we want to throw away bootstrap data, and save memory
gwas_link_parallel	if user wants to run in gwas linking in parallel, registerDoParallel must have been run in advance.
bootstrap_parallel	if user wants to run in bootstraps in parallel, registerDoParallel must have been run in advance.
keep_model	if we should keep the regression model, can be large.
keep_gwas	set to T if we want to include gwas in returned rolypoly object.
...	other arguments to pass to cv.glmnet

Value

rolypoly object

Examples

## Not run: rolypoly_roll(rolypoly)

Example output

Welcome to rolypoly! Say hi @diegoisworking.

rolypoly documentation built on May 2, 2019, 2:47 p.m.