dist_mat: Create Distance Matrix

In dswatson/bioplotr: Pretty, simple, optionally interactive plots for bioinformatics analysis pipelines

Create Distance Matrix

Description

This utility function calculates distance based on a user defined measure and optionally filters probes by leading log fold change.

Usage

dist_mat(dat, dist, p, top, filter_method, center, robust = FALSE)

Arguments

dat	Omic data matrix.
dist	Distance measure to be used.
p	Power of the Minkowski distance.
top	Optional number (if > 1) or proportion (if < 1) of top probes to be used for distance calculations.
filter_method	String specifying whether to apply a "pairwise" or "common" filter if top is non-NULL.
center	Center each probe prior to computing distances?
robust	Use robust probe centering?

Details

Data are optionally centered by probe and samplewise distance calculated using one of the following methods:

• "euclidean", "maximum", "manhattan", and "minkowski" are all documented in the dist function. bioplotr relies on a lower level implementation via Rfast::Dist to speed up computations.

• "cosine" and "canberra" are implemented and documented in wordspace::dist.matrix.

• "pearson", "kendall", and "spearman" correspond to various forms of correlation distance, generally defined as 1 – the absolute value of the correlation coefficient. See cor for more details.

• "bray", "kulczynski", "jaccard", "gower", "altGower", "morisita", "horn", "mountford", "raup", "binomial", "chao", "cao", and "mahalanobis" are all available and documented in the vegan::vegdist function. These are designed for use with ecological data, e.g. a matrix of microbial OTU counts.

• "bhattacharyya", "hellinger", "kullback_leibler", and "MI" are information theoretic distance metrics. The former three are implemented in Rfast::Dist. See bioDist::MIdist for details on the latter.

dswatson/bioplotr documentation built on March 3, 2023, 9:43 p.m.