Fig. 2

Strain-level phylogenetic analysis of the two most prevalent bacterial species identified in the metagenomic samples of the New York urban environment. The phylogenetic trees are inferred by applying StrainPhlAn on the raw sequencing reads. a Maximum likelihood phylogeny of P. stutzeri (built with RAxML [83] internally in StrainPhlAn). The root of the phylogenetic tree is placed using P. putida as an outgroup. Nodes are colored by the New York Borough from which the sample was collected, with black stars representing reference genomes. The height of the bars of the bar-plot on the outer ring represents the relative abundance of P. stutzeri as computed by MetaPhlAn2, while the color represents the surface material of the sample. The lengths of branches marked with a double horizontal line are reduced by 50% (b, c) PCA plot based on the genetic distance computed on the species-specific markers sequences of 416 samples and 18 reference genomes of P. stutzeri (b) and 111 samples and 80 reference genomes of S. maltophilia (c). The points are colored according to the New York Borough