mbecModelVariancePVCA: Estimate Explained Variance with Principal Variance Component...

In buschlab/MBECS: Evaluation and correction of batch effects in microbiome data-sets

Description

The function offers a selection of methods/algorithms to estimate the proportion of variance that can be attributed to covariates of interest. This shows, how much variation is explained by the treatment effect, which proportion is introduced by processing in batches and the leftover variance, i.e., residuals that are not currently explained. Covariates of interest (CoI) are selected by the user and the function will incorporate them into the model.

Usage

mbecModelVariancePVCA(
  model.vars,
  tmp.cnts,
  tmp.meta,
  type,
  pct_threshold,
  na.action
)

Arguments

model.vars	Covariates to use for model building.
tmp.cnts	Abundance matrix in 'sample x feature' orientation.
tmp.meta	Covariate table that contains at least the used variables.
type	String the denotes data source, i.e., one of "otu","clr" or "tss" for the transformed counts or the label of the batch corrected count-matrix.
pct_threshold	Cutoff value for accumulated variance in principal components.
na.action	Set NA handling, will take global option if not supplied.

Details

Principal Variance Component Analysis (pvca): Algorithm - calculate the correlation of the fxs count-matrix - from there extract the eigenvectors and eigenvalues and calculate the proportion of explained variance per eigenvector (i.e. principal component) by dividing the eigenvalues by the sum of eigenvalues. Now select as many PCs as required to fill a chosen quota for the total proportion of explained variance. Iterate over all PCs and fit a linear mixed model that contains all covariates as random effect and all unique interactions between two covariates. Compute variance covariance components form the resulting model –> From there we get the Variance that each covariate(variable) contributes to this particular PC. Then just standardize variance by dividing it through the sum of variance for that model. Scale each PCs results by the proportion this PC accounted for in the first place. And then do it again by dividing it through the total amount of explained variance, i.e. the cutoff to select the number of PCs to take (obviously not the cutoff but rather the actual values for the selected PCs). Finally take the average over each random variable and interaction term and display in a nice plot.

Value

Data.frame that contains proportions of variance for given covariates in a principal variance component analysis approach.

buschlab/MBECS documentation built on Jan. 21, 2022, 1:27 a.m.