dICC: Distance-based Intra-Class Correlation Coefficient

In GUniFrac: Generalized UniFrac Distances, Distance-Based Multivariate Methods and Feature-Based Univariate Methods for Microbiome Data Analysis

Distance-based Intra-Class Correlation Coefficient

Description

Distance-based Intra-Class Correlation Coefficient (ICC) is an extension of the traditional univariate ICC to multivariate case, where the relationship between the multivariate measurements is summarized in a distance matrix. It quantifies the ability of a measurement method in reproducing the inter-sample relationship.

Usage

 dICC(dist.mat, strata)

Arguments

dist.mat	a symmetrical distance matrix between all the replicates (technical and biological replicates).
strata	a factor with each level corresponding to a biological replicate.

Value

Function dICC returns a list with the following component:

ICC	the distance-based ICC value.

Author(s)

Jun Chen and Xianyang Zhang

References

Chen, J. & Zhang, X. 2022. dICC: Distance-based Intraclass Correlation Coefficient for Metagenomic Reproducibility Studies. submitted.

See Also

dICC.SE.asympt, dICC.SE.bt

Examples

# Generate the error-free measurements of 20 biological replicates, each with four dimensions
y <- matrix(rnorm(80), nrow = 20, ncol = 4)

# Generate two technical replicates for each biological replicate and add measurement errors
y1 <- y + matrix(rnorm(80, sd = 0.5), nrow = 20, ncol = 4)
y2 <- y + matrix(rnorm(80, sd = 0.5), nrow = 20, ncol = 4)
y12 <- rbind(y1, y2)

# Create the design vector
strata <- factor(rep(1 : 20, 2))

# Calculate the distance and distance-based ICC
dist.mat <- as.matrix(dist(y12))
dICC(dist.mat, strata)

GUniFrac documentation built on Sept. 14, 2023, 1:07 a.m.