Extended Data Fig. 4: Phylogenetic tree based on the gp17 sequences of 26 phage clones from the three-host treatment (EPEC, MG1655, REPEC).

Sequences were aligned with MUSCLE (v3.8.31) in MEGA X^51,52. The phylogenetic tree was reconstructed using the maximum likelihood method as implemented in PhyML (v3.1/3.0)^53,54. The WAG substitution model⁵⁵ was selected assuming an estimated proportion of invariant sites (of 0.951) and 4 gamma-distributed rate categories to account for rate heterogeneity across sites. The gamma shape parameter was estimated directly from the data (gamma=91.073). Reliability of internal branches was assessed using the aLRT test (SH-Like). Final Log-Liklihood: −1839.18. Phage clones labelled by phenotype and day of generation. ‘S’ or ‘G’ refers to specialist or generalist according to phenotypic measurements (Figs. 3–4) and the number refers to the day of isolation. C8, D8, C9, D9, E8, F8 refer to the specific three-host experimental population from which the phage clone was isolated. Each phage clone was obtained by plating the whole population lysate onto an individual host (Figs. 3–4). The strain name refers to that host of isolation. For example, ‘G10 C8 MG1655’ refers to a phage clone with a generalist phenotype isolated from a day 10 population C8, isolated from a plaque on MG1655. Numbers on branches show bootstrapping support for that branch (percent, 100 bootstraps). Strains are clustered based on those clades with greater than 80% support. Dashed lines are for labels and are not part of the tree.