Fig. 5.

Heat map showing the 50 most abundant antimicrobial resistance genes (ARGs) by mean relative abundance. ARG (x-axis) relative abundances in RPKM have been log₁₀-transformed and clustered by specific features of the samples (y-axis) including sample age, E. coli relative abundance, Proteobacteria relative abundance, delivery mode, and the number of unique ARGs per sample. ARGs are clustered by the Euclidean distance and samples are clustered using the Canberra distance. Hierarchical clustering of samples was used to determine 8 clusters as indicated by the numbers to the right of the heat map