computeBCeF2: Generates values for roc curve for evaluating the performance...

In NabilaRahman/batchPred: batchPred - Objective prediction of relevant set of batch effects variables

Generates values for roc curve for evaluating the performance of batch correction

Description

This package corrects for confounders in gene expression datasets using multiple linear regression model and then evaluates the improvement in gene coexpression of using a gold standard co-expression network.

Usage

computeBCeF2(
  input.edata,
  covariates.df.list,
  input.gold.standard,
  plot.title = "Batch Correction Evaluation",
  roc.curve.legend = NULL,
  line.color = NULL
)

Arguments

input.edata	Matrix of raw expression dataset to be adjusted. Rows represent the genes/probes and columns represent samples.
covariates.df.list	List of covariate dataframes. Each dataframe in the list consists of columns representing known covariates e.g. Age, Sex, or unknown covariates such as principle components. Rows represent samples.
input.gold.standard	Dataframe. A gold standard that includes a gene coexpression confidence. First column represents the first gene, the second column the second gene and the third columns should be a binary vector where 1 indicates true associations and 0 indicates false associations. The gene IDs in the first two column of input.gold.standard and the rownames of input.edata must be of the same type.
plot.title	Title of the plot.
roc.curve.legend	Character vector of descriptions for each set of covariates to plot. Must be in the same order as the dataframes in covariates.df.list. This modifies the roc curve legend.
line.color	Character vector specifying color of roc curves. By default sets color for upto 6 roc curves with raw displayed as black. Manually specify line.color if plotting more than 6 different sets of covariates, i.e. over 6 dataframes in covariates.df.list.

Value

computeBCeF2 Generates a list of all variables needed for plotting roc curves of raw and batch corrected datasets. One batch corrected roc curve is computed for each dataframe specified in covariates.df.list, allowing you to compare the performance of different sets of covariates. Use plotBCeF2 to plot these variables

NabilaRahman/batchPred documentation built on June 19, 2022, 5:35 a.m.