get_linear_model: Train a linear model

In CoRegFlux: CoRegFlux

Description

Here we train a linear regression model of the form x= alpha + beta*I where x is the gene expression of the metabolic genes of the train data set train_expression, alpha is an intercept, I is the influence of the regulators of the training data set and beta are the coefficients.

Usage

get_linear_model(train_expression,
  train_influence = regulatorInfluence(network, train_expression, minTarg
  = 10), network)

Arguments

train_expression	Gene expression of the training data set, not necessary if train_influence is supplied. Should be numerical matrix corresponding to the gene expression. Rownames should contain gene names/ids while samples should be in columns.
train_influence	Optional. Regulator influence scores computed using the function CoRegNet::regulatorInfluence for the training data set, default minTarg = 10
network	CoRegNet object use to build the linear model and to compute the influence.

Details

train_expression Gene expression of the training data set, not necessary if train_influence is supplied. Should be numerical matrix corresponding to the gene expression. Rownames should contain gene names/ids while samples should be in columns.

Value

A linear model

See Also

predict_linear_model_influence

CoRegFlux documentation built on Nov. 8, 2020, 5:21 p.m.