COVID-19 respiratory microbiome analysis code

Code accompanying the manuscript "Clinical practices underlie COVID-19 patient respiratory microbiome composition and its interactions with the host". Link to preprint: https://www.medrxiv.org/content/10.1101/2020.12.23.20248425v3

Contributors:

• Verónica Lloréns Rico, PhD
• Ann C. Gregory, PhD

File description:

Code files:

• script_covid_preprocessing.R: raw data preprocessing
• script_covid_dataexploration.R: initial data exploration and plots
• script_covid_alphadiv.R: Alpha diversity analyses and glmm modeling
• script_covid_betadiv.R: dbRDA analyses of microbiome composition
• script_covid_diffabundances.R: differential taxon abundance analyses
• script_covid_species_strain_analyses.R: macro-vs-micro diversity analyses
• script_covid_sc_analyses.R: analyses on the scRNA-seq data of the lower respiratory tract cohort

Additional folders:

• data/: use this folder to download the raw data and metadata tables from EGA (with controlled access, accession number EGAS00001004951). Contains two additional files:
  • sample_ids.txt: contains sample IDs to preprocess the data (used in script_covid_dataexploration.R)
  • coding_table_final.txt: contains explanations of the metadata variables (used in script_covid_betadiv.R)
• R/: additional R functions, called by the different scripts

Installation instructions

Hardware required

All code was run on laptop/desktop computers with 8 cores with 16GB RAM

Software required

Required: R (https://cran.r-project.org/). The code in this repository was run on R v.4.0.5. Recommended: Rstudio (https://www.rstudio.com/)

R packages required:

• dada2 (v1.18.0)
• phyloseq (v1.34.0)
• ggplot2 (v3.3.3)
• ggpubr (v0.4.0)
• cowplot (v1.1.1)
• tidyverse (v1.3.1)
• reshape2 (v1.4.4)
• compositions (v2.0.1)
• vegan (v2.5.7)
• rstatix (v0.7.0)
• ALDEX2 (v1.22.0)
• ggrepel (v0.9.1)
• tibble (v3.1.2)
• DECIPHER (v2.18.1)
• Biostrings (v2.58.0)
• biomod2 (v3.5.1)
• wesanderson (v0.3.6)
• colortools (v0.1.5)
• ggiraph (v0.7.10)
• ggiiraphExtra (v0.3.0)
• glmulti (v1.0.8)
• sjPlot (v2.8.7)
• lme4 (v1.1.27)
• CoDaSeq (v0.99.6)
• DESeq2 (v1.30.1)
• mixOmics (v6.14.1)
• lubridate (v1.7.10)
• Seurat (v4.0.4)
• chisq.posthoc.test (v0.1.2)
• ggmosaic (v0.3.3)
• patchwork (v1.1.1)
• gtable(v0.3.0)

Running instructions

1. Download raw sequencing files and metadata from EGA: link to study
2. Store them in the data/ folder, or any other directory of your choice.
3. Run script_covid_preprocessing.R. This will do all the raw data preprocessing: quality control, trimming, denoising, ASV assignation and taxonomic annotation, as well as decontamination.
4. Run script_covid_dataexploration.R. This will do the initial data exploration, and generate some of the plots available in Figure 1.
5. Run script_covid_alphadiv.R. This will perform the alpha diversity analyses and modeling section.
6. Run script_covid_betadiv.R. This will perform the beta diversity analyses and modeling section.
7. Run script_covid_diffabundances.R. This will perform the differential taxon abundances reported in the manuscript.
8. Run script_covid_species_strain_analyses.R. This will perform the analyses on macro- and micro-diversity, shown in Figure 2.
9. Download and process raw sequencing files from the scRNA-seq lower respiratory tract cohort. link to study and link to study website
10. Run script_covid_sc_analyses.R. This will perform the analyses on the lower respiratory tract cohort, shown in Figure 3.