These files are relevant to the study https://www.biorxiv.org/content/10.1101/2025.03.20.644349v2 (accepted for publication in Nature Communications) and the associated code available at https://github.com/Baud-lab/P50/tree/Master/16S and https://github.com/Baud-lab/CoreQuantGen. The folder source_data_paper includes the source data that can be used to reproduce the figures in the paper.175568_57950_analysis_16S_FilterfeaturesagainstreferencefilterfeaturesPhylogenetictreedatabasesgg202210202210taxonomyasvnwkqzaBIOM file from Qiita (https://qiita.ucsd.edu/) artefact 175568 in analysis 57950, created by merging artefacts 175546, 175547, 175548, 175549, 175550, 175551, 175552, 175553, and 175555 (corresponding to the different library preparations) from study 11479.Includes abundance and taxonomy for all ASVs identified in the samplestaxonomy_Greengenes2.txt1) downloaded 2022.10.taxonomy.asv.nwk.qza from http://greengenes.microbio.me/greengenes_release/2022.10-rc1/;2) downloaded file 175568_feature-table.qza from Qiita (https://qiita.ucsd.edu/) artefact 175568 in analysis 57950 (same information as BIOM table above)3) filtered the taxonomy file to keep only the ASVs present in the HS samples using qiime greengenes2 taxonomy-from-table --i-reference-taxonomy 022.10.taxonomy.asv.nwk.qza --i-table 175568_feature-table.qza --o-classification biom.taxonomy.qza4) exported taxonomy to TSV file qiime tools export --input-path biom.taxonomy.qza --output-path taxonomy.tsv5) renamed taxonomy as taxonomy_Greengenes2.txtCol1 is ASV sequennceCol2 is full taxonomyCol3 and Col4 not used.full_biomt_clr_counts.RDataContains "full_biomt" and "clr_counts" (each for 3886 rats and 93,090 ASVs)deblur_rarefied_collapsed_full_biomt.RDataR object corresponding to Qiita artefact ID 214605. Contains "collapsed_full_biomt" (subset of 2,728 rats with >10,000 sequencing reads and 2,251 taxa from phylum to species)collapsed_full_biomt_collapsed_clr_counts.RDataContains "collapsed_full_biomt" and "collapsed_clr_counts" (created by 4_merge_taxonomic_level.R on the associated github). 3,886 rats and 2,675 taxa (phylum to species).metadata_16Spaper.RDataAll metadata relevant to the 16S dataaugmented_VC.RDataoutput of CoreQuantGen code, for microbiome phenotypes and using a model with DGE only (no IGE), augmented using annotate_VCs_pvalues.RColumns trait2, sample_size2, sample_size2_cm None, study2 all None in univariate analysisOther columns:- trait1 phenotype name (ASV or taxon _ cohort)- sample_size1 number of individuals with phenotype, covariate, cage and genetic information- sample_size1_cm (equal or greater than sample_size1): includes cage mates of individuals in sample_size1 for which phenotype and/or covariate is NA but with cage and genetic information- covariates_names for all phenotypes, residuals from regressing out the covariates were used. Hence, only a mean term (showing as m,e,a,n) and a group size term (not showing) were fit as fixed effect in the linear mixed models used for all genetic analyses.- conv indicates if the model converged (True)- LML -log10(Maximum Likelihood)- prop_Ad1 proportion of phenotypic variance (total_var1) explained by host genetic effects/DGE- prop_Ed1 proportion of phenotypic variance (total_var1) explained by individual environmental effects- prop_Dm1 proportion of phenotypic variance (total_var1) explained by maternal effects- prop_C1 proportion of phenotypic variance (total_var1) explained by cage effects- tot_genVar1 proportion of phenotypic variance (total_var1) explained by DGE (and IGE and DGE-IGE covariance- total_var1 total phenotypic variance- id.x not used- time_exec time it took to fit the model (and estimate STE)- STE_Ad1 standard error for prop_Ad1- STE_Ed1 standard error for prop_Ed1- STE_Dm1 standard error for prop_Dm1- STE_totv1 standard error for total_var1- corParams_Ed1_Dm1 correlation between the parameters prop_Ed1 and prop_Dm1- id.y not used- taxon1 taxon of trait1- study1 cohort of trait1- full_taxon full taxonomy for taxon1- pvalue_DGE p-value for H0: prop_Ad1 = 0- sw_bonf_pvalue_DGE and sw_qvalue_DGE not used- cw_bonf_pvalue_DGE Bonferroni corrected p-value accounting for the number of phenotypes considered in a given cohort- cw_qvalue_DGE Q value accounting for the phenotypes considered in a given cohortphenos_all_estNste.RDatasimilar to augmented_VC.RData but for organismal phenotypes measured in HS rats.all_VCs_corr_Ad1d2_zero_P50_Rn7_pruned_DGE.RDatasimilar to augmented_VC.RData but from a bivariate modeltrait2 is same taxon or ASV as trait1 but a different cohortcorr_Ad1d2 is the (genetic) correlation between DGE on trait1 and DGE on trait2augmented_IGE_VC.RDatasimilar to augmented_VC.RData but from a model including DGE and IGE. whole sample data only (not per cohort).microbiome_DGE_QTLs.RDataMicrobiome-associated loci (for ASVs and taxa, -logP > 4)microbiome_DGE_QTLs_ALL.RDataMicrobiome-associated loci (for ASVs and taxa, -logP > 4)P50_Rn7_chr10qtl_allSNPS.rawGenotypes for SNPs at the chromosome 10 replicated locus.permutations_cagemates folderFor Fig 6c and SFig 13. Contains output of variance decomposition after permuting the cage mates and analysing using three different models. to be explored with analyse_scrambled.R in that same folder.MI and NY foldersSimulations for Fig. 6D and Supplementary Fig. 14. Each folder includes simulated values and estimated values from the simulations presented in the paper and more. 0.9 0 and neg0.9 refer to the (genetic) correlation between DGE and IGE simulated. 21 and 22 are the seeds used for the simulationsIGE_null_simulations_SFig12For SFig. 12. Contains output of variance decomposition for null (no IGE) simulations and analysing using three different models. QQ plots from 2_pvalues_VCs_bootstrap.R in that same folder.

16S metabarcoding databases and naive-bayes classifiers specific to the V4-V5 region. Built from the Silva 138.1 SSU Ref NR 99 database using Qiime2 (version 2023.2 and 2023.5) and the q2-clawback plugin. Includes weighted classifiers for two Earth Microbiome Project Ontology (EMPO) 3 habitat types: "sediment (saline)" and "water (saline)" , with data downloaded from Qiita. Sequences were not dereplicated.

Primers used:

EMP 16S 515f: GTGYCAGCMGCCGCGGTAA

EMP 16S 926r: CCGYCAATTYMTTTRAGTTT

Stats

286,948 unique sequences

388,496 total sequences

46,254 unique taxa (Level 7)