A global survey of host, aquatic, and soil microbiomes reveals shared abundance and genomic features between bacterial and fungal generalists

This repository provides the code required to reproduce the analyses of the following publication:

Daniel Loos, Ailton Pereira da Costa Filho, Bas E. Dutilh, Amelia E. Barber, Gianni Panagiotou. A global survey of host, aquatic, and soil microbiomes reveals shared abundance and genomic features between bacterial and fungal generalists. Cell Reports. 2024. https://doi.org/10.1101/2022.11.15.515575.

Correspondence: [email protected] or [email protected]

Abstract

Environmental change, coupled with alteration in human lifestyles, is profoundly impacting the microbial communities critical to the health of the Earth and its inhabitants. To identify bacteria and fungi that are resistant and susceptible to habitat change, we analyze thousands of genera detected in 1,580 host, soil, and aquatic samples. This large-scale analysis identifies 48 bacterial and 4 fungal genera that are abundant across the three biomes, demonstrating fitness in diverse environmental conditions. Samples containing these generalists have significantly higher alpha diversity. These generalists play a significant role in shaping cross-kingdom community structure, boasting larger genomes with more secondary metabolism and antimicrobial resistance genes. Conversely, 30 bacterial and 19 fungal genera are only found in a single habitat, suggesting a limited ability to adapt to different and changing environments. These findings contribute to our understanding of microbial niche breadth and its consequences for global biodiversity loss.

Get Started

Source code is stored in the directory src. This workflow starts with a nextflow pipeline to retrieve taxon abundance profiles in directory src/its-16s-nf. Then, a drake pipeline is used for subsequent statistical analysis and visualization in directory src/plans. The main entry point for this analysis is the Makefile providing shell commands make build for building the docker image and make analyse to execute the pipeline. Briefly, R packages were installed using script src/install/install.R. Then script main.R is executed that bootstraps the global R environment using script src/defaults.R and executes the individual drake targets defined in directory src/plans. The methods section of the paper provides further details.

Notable targets of the pipeline

Target name	Description
bioprojects	Cohorts used in this meta study
samples	Table of biological samples with meta data
abundances	Abundance profiles of all samples
alphadiv	Alpha diversities per sample subset
stepwise_constrained_ordinations	dbRDA to ordinate the abundance profile of one kingdom using the best explaining taxa from the other kingdom
selected_generalists_specialists	Generalists and specialists of the meta study
generalists_common_coabundance_graph	Co-abundances common in most environments
fig6_heatmap	Abundance heatmap of the paper

Supplementary analysis

The get started section describe analysis done with R Drake. Here we describe extra analysis performed with R Targets and Snakemake. They are available under src/sup. They include:

Task	Description
Primer detection	16S and ITS primers for bioprojects without publication were discovered based on alignment
Generalist and specialists BGCs	Fungal and Bacterial BGCs were identified with antiSMASH
Bacterial generalists and specialists AMR	Antimicrobial Resistance Genes identification
Decontamination	Perform decontamination of low biomass samples for 16S and ITS