compute_term_pcs: Compute principal components of SNPs annotated to ontology...

In SimonLarsen/higana: Ontology-based analysis of genetic variants.

Description

Compute principal components of SNPs annotated to ontology terms.

Usage

compute_term_pcs(
  path,
  o,
  data,
  terms = NULL,
  stand = "binom2",
  max_term_size = 75,
  min_term_size = 3,
  explain_var = 0.95,
  max_pcs = 50,
  num_parts = 1,
  exclude_snps = NULL,
  rsvd_threshold = Inf
)

Arguments

path	Path to output file.
o	An annotated ontology object.
data	A BEDData object
terms	A character vector of terms to compute PCs for. Will use all terms if not provided.
stand	Which standardization method to use. One of "none", "binom" (old Eigenstrat-style), "binom2" (new Eigenstrat-style), "sd" (zero-mean unit-variance) or "center" (zero mean).
max_term_size	Skip terms annotated with fewer than this number of genes.
explain_var	Restrict number of PCs to explain at least this fraction of variance.
max_pcs	Return at most this number of PCs.
num_parts	Number of parts to write output to.
exclude_snps	Named list of SNP IDs to exclude from terms.
rsvd_threshold	Use randomized SVD when number of variables exceeds this threshold. Set to Inf to turn off.

SimonLarsen/higana documentation built on April 24, 2020, 7:29 a.m.