Extended Data Fig. 4: Functional ordination of bacterial genomes from the SMGA and genome-inferred functional traits (GIFTs) abundance.

a) Ordination using a t-Distributed Stochastic Neighbor Embedding (t-SNE) analysis for clustering GIFTs values of the SMGA’s bacterial genomes. GIFTs values indicate functional capabilities of bacteria based on their genome content. These metrics are computed for individual functions within a genome, representing the proportion of biochemical reactions mediated by the genes to fulfill specific metabolic tasks. Genomes are colored by taxonomic Order. b) Identical ordination but genomes colored by their Metabolic Capacity Index (MIC). Color gradient indicates the increase of MIC; increasing MIC can be observed from bottom to top. Low-MCI: MCI < 0.15; mid-MCI: 0.15 < MCI < 0.3; high-MCI: MCI > 0.3. A PERMANOVA (Permutational Multivariate Analysis of Variance) was conducted to analyse differences in the MCI across genomes of bacteria collected from fish in fresh and seawater. PERMANOVA results are reported under the MCI legend, ** indicates a significance of p < 0.01. The PERMANOVA was two-sided, and no adjustments were made for multiple comparisons. c) Heatmap showing the abundance of GIFTs associated with a particular metabolic function (columns) in each bacterial genome from the SMGA (rows). Bacterial genomes were hierarchically clustered according to their GIFTs abundance (left dendrogram).