psiform: A tail-based test statistics (TTS) for differential...

In SkadiEye/Psiform: Shared Informative Factor Models for integration of multi-platform Bioinformatic data with Pathway information incorporation

Description

A tail-based test statistics (TTS) for differential expression detection.

Usage

psiform(
  ...,
  y,
  G = NULL,
  dim_u = 5,
  weight = "none",
  scale = FALSE,
  lambda = NULL,
  a = 3.7,
  param_init = NULL,
  n_iter = 100,
  epsilon = 10^-8,
  conv_thresh = 10^-6,
  method = 1
)

Arguments

...	One or multiple matrices of expression levels from multiple platforms.
y	Subject labels.
G	Pathway information.
dim_u	Number of dimension of u.
weight	Type of weight on the pathway. Possible values include "sqrt-degree", "none" or "degree". The default is "none".
scale	Whether to scale input matrices.
lambda	Hyper-parameters in SCAD, TLP, and L1 of phenotype differences.
a	Hyper-parameter in SCAD. Default is 3.7.
param_init	If not NULL, should be a list including c, beta, u, v and sigma_sq.
n_iter	Number of maximum iterations.
epsilon	A small number to avoid zero-denominators.
conv_thresh	Threshold for convergence.
method	Type of TLP.

Value

Returns a Psiform object.

SkadiEye/Psiform documentation built on May 15, 2020, 12:13 a.m.