This respository contains R code associated with "Microbiome stability and structure is governed by host phylogeny over diet and geography in woodrats (Neotoma spp.)"
16s_processing.R: Initial processing of bacterial 16s amplicon sequences, intitial filtering, and quality control. Amplicon sequences are available from the NCBI sequence archive under BioProject PRJNA722312. Inputs are fastq.gz files from each sequencing run. Outputs are rarefied and non-rarefied microbiome phyloseq objects for downstream analyses.
trnl_processing.R: Initial processing of plant trnL amplicon sequences. Amplicon sequences are available from the NCBI sequence archive under BioProject PRJNA722312. This script is very similar to the 16s_processing.R file, with modifications for different primers and different taxonomy assignment methods. Taxonomy is assigned using separate custom python script: see https://github.com/robertgreenhalgh/stand. Inputs are fastq.gz files from each sequencing run. Output is a plant ASV table.
Simmr_d13C_models.R: Script for running stable isotope mixing models of δ13C values from woodrat fecal samples to estimate dietary proportions of C3 vs C4/CAM plants.
cactus_adjust.R: Adjust sequencing based diet data to account for missing cactus reads using stable isotope data. Script creates a phyloseq object from trnl data, filters ASVs, and creates basic plots demonstrating missing cactus in diets. Individual diets are adjusted using carbon stable isotope data based on SIMMR models from captive cactus diet trials and wild collected fecal samples. Produces the cactus adjusted, filtered, rarefied plant OTU table used in downstream analyses.
Diversity_Composition_comps.R: Script for alpha and beta diversity analyses for wild rat microbiome and diet data. Includes code for Figure 2.
MRM_models.R: Script for multiple regression on distance matrices (MRM) models and code for Figure 3. Code for Figure 4B is also based on this script.
Differential_Abundance_diet.R: Script for differential abundance analyses identifying ASVs (or bacterial families) associated with different wild diet components using DESeq2.
Sympatry.R: Script for testing whether heterospecifics from the same site had more similar microbiomes (based on Jaccard similarity) than matched species pairs from different sites. Script also includes code to compare average percent of shared ASVs for sympatric and equivalent allopatric populations, and code for Figure S5.
mantel_and_clustering.R: Code for two approaches for examining correlations between the microbiome and geography, phylogeny, or diet. Includes comparisons at the individual, population, and species levels. Mantel and Partial Mantel tests are based on distance matrices. Hierarchical clustering analyses compare tree congruence, calculating significant congruence via a bootstrapped p-value.
Differential_Abundance_captivity.R: Code for differential abundance analyses identifying ASVs (genera, and families) that increased/decrease in wild v. captive animals. Includes analyses of taxa that change in each population and when all individual are grouped together. Script is very similar to Differential_Abundance_diet.R but includes loop to run through all populations.
Captivity_impacts.R: Exploring how captivity impacts microbiome diversity and composition. Includes PERMANOVAs, builds a dataframe of changes in observed richness and composition for each animal, and includes statistical analyses using that dataframe.
Morans.R: Testing for phylogenetic signal in how animal microbiomes change in captivity. Code uses the dataframe produced in Captivity_impacts.R.
rare_lost.R: Using paired wild and captive 16s data to test whether ASVs that are less abundant in wild animals are more likely to be lost in captivity.
Captivity_homogenize.R: Code to test whether captivity reduces variation among individuals.
PNM_sites_taxa.R: Code for fitting the prokaryote neutral model (PNM) to ASV data, testing whether model fit correlates with host species diversity at a site, calculating model fit for wild and captive animals, and identifying "neutral" and "non-neutral" taxa. The code for the PNM comes from the supplemental material (Supplemental Code 1) from Burns, A., Stephens, W., Stagaman, K. et al. Contribution of neutral processes to the assembly of gut microbial communities in the zebrafish over host development. ISME J 10, 655–664 (2016). https://doi.org/10.1038/ismej.2015.142.