Nasopharyngeal and oropharyngeal microbiome of rural residents in the Netherlands: paired 16S rRNA gene sequencing data from two timepoints (2014–2015 and 2021–2022)

This dataset consists of three preprocessed phyloseq S4 objects, each containing an amplicon sequence variant (ASV) table, a taxonomy table, and sample metadata. Metadata has been limited to retain participant confidentiality. Additional metadata may be made available upon reasonable request.

The three objects are:

VGO3_ps_crosssectional — cross-sectional data from T1 (2021–2022), used for cross-sectional associations between URT microbiome composition and lung function

VGO3_ps_longitudinal_stability — paired T0 (2014–2015) and T1 samples from participants with microbiome data at both timepoints in the same niche, used for within-person microbiome stability analyses

VGO3_ps_longitudinal_lungfunction — T0 microbiome data from participants with lung function measurements at both timepoints, used for longitudinal lung function prediction analyses

Sample collection and sequencing

Nasopharyngeal (NP) and oropharyngeal (OP) swabs were collected during home visits by trained field staff using Copan ESwabs, transported at approximately 4°C to the Institute for Risk Assessment Sciences (IRAS), and stored at −80°C until DNA extraction and sequencing. Sequencing protocols followed those used in the original cohort studies. Baseline NP samples and all follow-up samples targeted the V4 hypervariable region of the bacterial 16S rRNA gene, whereas baseline OP samples were originally sequenced targeting the V3–V4 region. To ensure comparability across datasets, V3–V4 reads were trimmed to the V4 region using cutadapt and harmonised to the 515F/806R primer pair; this harmonisation step did not materially alter overall microbiome composition. Paired-end reads were processed using the DADA2 pipeline (v1.16.0; Callahan et al., 2016) to infer ASVs directly from quality-filtered reads.

Preprocessing

All three phyloseq objects have undergone the following preprocessing steps prior to analysis: removal of contaminant ASVs using the decontam package combined with manual inspection and rescue of known commensal taxa, exclusion of samples with bacterial DNA concentration below 0.095 pg/µL (qPCR-based), exclusion of samples with fewer than 10,000 reads, and prevalence- and abundance-based taxon filtering. Full details of all preprocessing steps are available in the analysis scripts linked below.

Code availability

Analysis scripts are available at: https://github.com/BeatriceCornuHewitt/Microbiome_lungfunction.git

Specifically, preprocessing steps for the cross-sectional object are documented in 2_Cross_sectional_analyses.Rmd and for the longitudinal objects in 3_Longitudinal_analyses.Rmd.