2.8 Statistical analysis for β‐diversity of gut bacteria between population and differential abundance
Weighted and unweighted UniFrac distance matrix between samples were used to investigate differences in gut microbial communities between E-mink population and host sex. These metrics take into account the differences in phylogenetic distance and abundance of each bacterial community between samples, pairwise, emphasizing on rare (unweighted) or abundant microbial taxa (weighted). PERMANOVA tests were conducted in similar conditions than PERMANOVA tests conducted on MHC matrixes. Additionally, models were constructed with adaptive genetic variables: motif richness and divergence from both MHC genes. To test the hypothesis that both neutral and adaptive distances are correlated with gut microbial composition, we employed Mantel tests between each genetic distance and Unifrac distance.
A differential abundance analysis was conducted on the raw ASVs count (after filtering, prior to rarefaction) that were present in more than 10% of all the samples and that had a relative abundance of more than 5% among all taxa. It corresponds to the core microbiota of the dataset, represented by 1203 phylotypes. The abundance analysis was made at the ASV level with the DESeq2 package, using a negative binomial Wald test to test significance in contrast between each E-mink population and each common MHC motifs that were present in at least two individuals. Only microbial ASV with a significance level (α) below .001 after false discovery rate (FDR) corrections were considered using the Benjamini–Hochberg method (Love et al., 2014). From the same core microbiota, we also tested for Pearson correlations between microbial genera’ relative abundance per individual and continuous genetic variables (MLH, motif richness and PD for both MHC class I and II) with FDR corrections, for genera that were encountered at least in three individuals.