learn_covariance: Function to learn covariance structure across samples using...

In davidaknowles/suez: Molecular response QTL mapping

Description

Function to learn covariance structure across samples using latent factors

Usage

learn_covariance(
  input,
  anno,
  kinship_matrix = NULL,
  P = 20,
  same_ind_initial_weight = 0.01,
  kinship_matrix_initial_weight = 0.01,
  ...
)

Arguments

input	[genes x samples] Normalized (e.g. quantile normalized) matrix of log2 gene expression
anno	[samples x 2] data.frame with 'individual' and 'condition' columns. If a kinship matrix is provided the row and column names must correspond to the values used in the 'individual' column.
kinship_matrix	Representing relatedness between individuals. Optional.
P	Number of latent factors to model. suez can prune out unnecessary factors (in principle).
same_ind_initial_weight	Initialization.
kinship_matrix_initial_weight	Initialization.
...	passed to rstan::optimizing

davidaknowles/suez documentation built on March 20, 2021, 2:19 a.m.