specificity: an R package for analysis of feature specificity to environmental and higher dimensional variables, applied to microbiome species data

Abstract

Understanding the factors that influence microbes’ environmental distributions is important for determining drivers of microbial community composition. Species distributions are governed by parameters that influence their dispersal or survival. These include environmental variables like temperature and pH, and higher-dimensional variables like geographic distance and host species phylogenies. In microbial ecology, “specificity” is often described in the context of symbiotic or host parasitic interactions, but specificity can be more broadly used to describe the extent to which a species occupies a narrower range of an environmental variable than expected by chance. Using a standardization we describe here, Rao’s (1982, 2010) Quadratic Entropy can be conveniently applied to calculate specificity of a focal species to many different environmental variables, including 1-dimensional variables (vectors), dissimilarity matrices, phylogenies, and ontologies. We present our R package “specificity” for performing the above analyses, and apply it to four real-life microbial datasets to demonstrate its application: fungi living within the leaves of native Hawaiian plants, bacteria from the human gut microbiome, bacteria living within Antarctic glacier ice, and the Earth Microbiome Project data set. Finally, we present an interactive visualization and data exploration tool for these analyses, available as a companion R-package called “specificity.shiny”.

Specialty grand challenge: how can we use integrative approaches to understand microbial community dynamics?

“…Another way of imparting distinction is to tag microbial species as either specific (existing within a narrow range of a particular covariate) or cosmopolitan (existing in a broader range of a particular covariate). In this regard, R's Specificity Package (Darcy et al, 2022) is an important development (see Figure 8). It calculates Rao's Quadratic Entropy (RQE) as RQE s−1 i 1 s j i+1 D ij p i p j where microbial abundance p i p j is the multiplication of the abundance of a specific microbe in samples i and j, respectively, each weighted by the difference in the covariate value D ij .…”

Phylosymbiosis drives the distribution of pathogen-protective bacteria in Appalachian salamanders

“…To quantify the specificity of ASVs in the salamander skin microbiome to host phylogeny and environment, we used the Rao's Quadradic Entropy 56,57 method implemented in the specificity R package 58 . We calculated specificity to four distance variables used in the MRM analysis.…”

Temporal stability and community assembly mechanisms in healthy broiler cecum

“…These rare taxa provide a reservoir of functions and offer a means toward achieving resilience. Beyond this approach, we apply a null modelingbased specificity measure by Darcy et al (2022) that is able to pick out taxa becoming active at narrow ranges. Our original study (Ijaz et al, 2018) has three diets, starter (days 0-10), grower (days 11-25), and finisher (days 26-35).…”