standardize_feature: Centralize (by design matrix) and standardize (by pooled...

In MMUPHin: Meta-analysis Methods with Uniform Pipeline for Heterogeneity in Microbiome Studies

Description

Centralize (by design matrix) and standardize (by pooled variance across all batches) feature abundances for empirical Bayes fit

Usage

standardize_feature(y, i_design, n_batch)

Arguments

y	vector of non-zero abundance of a single feature (if zero-inflated is true).
i_design	design matrix for the feature; samples with zeros are taken out (if zero-inflated is true).
n_batch	number of batches in the data.

Value

a list with component: y_stand for vector of centralized and standardized feature abundance, and stand_mean/varpooled for the location and scale factor (these are used later to back transform the batch-shrinked feature abundance).

MMUPHin documentation built on April 9, 2021, 6:01 p.m.