Inferring strain-level mutational drivers of phage-bacteria interaction phenotypes arising during coevolutionary dynamics

The enormous diversity of bacteriophages and their bacterial hosts
presents a significant challenge to predict which phages infect a focal
set of bacteria. Infection is largely determined by complementary – and
largely uncharacterized – genetics of adsorption, injection, cell
take-over and lysis. Here we present a machine learning approach to
predict phage-bacteria interactions trained on genome sequences of and
phenotypic interactions amongst 51 Escherichia coli strains and 45 phage l
strains that coevolved in laboratory conditions for 37 days. Leveraging
multiple inference strategies and without a priori knowledge of driver
mutations, this framework predicts both who infects whom and the
quantitative levels of infections across a suite of 2,295 potential
interactions. We found that the most effective approach inferred
interaction phenotypes from independent contributions from phage and
bacteria mutations, accurately predicting 86% of interactions while
reducing the relative error in the estimated strength of the infection
phenotype by 40%. Feature selection revealed key phage l and E. coli
mutations that have a significant influence on the outcome of
phage-bacteria interactions, corroborating sites previously known to
affect phage l infections, as well as identifying mutations in genes of
unknown function not previously shown to influence bacterial resistance.
The method's success in recapitulating strain-level infection
outcomes arising during coevolutionary dynamics may also help inform
generalized approaches for imputing genetic drivers of interaction
phenotypes in complex communities of phage and bacteria.