The emergence of microbial antibiotic resistance is a global health threat. In clinical settings, the key to controlling spread of resistant strains is accurate and rapid detection. As traditional culture-based methods are time consuming, genetic approaches have recently been developed for this task. The detection of antibiotic resistance is typically made by measuring a few known determinants previously identified from genome sequencing, and thus requires the prior knowledge of its biological mechanisms. To overcome this limitation, we employed machine learning models to predict resistance to 11 compounds across four classes of antibiotics from existing and novel whole genome sequences of 1936 E. coli strains. We considered a range of methods, and examined population structure, isolation year, gene content, and polymorphism information as predictors. Gradient boosted decision trees consistently outperformed alternative models with an average accuracy of 0.91 on held-out data (range 0.81–0.97). While the best models most frequently employed gene content, an average accuracy score of 0.79 could be obtained using population structure information alone. Single nucleotide variation data were less useful, and significantly improved prediction only for two antibiotics, including ciprofloxacin. These results demonstrate that antibiotic resistance in E. coli can be accurately predicted from whole genome sequences without a priori knowledge of mechanisms, and that both genomic and epidemiological data can be informative. This paves way to integrating machine learning approaches into diagnostic tools in the clinic. One of the major health threats of 21st century is emergence of antibiotic resistance. To manage its human health and economic impact, efforts are made to develop novel diagnostic tools that rapidly detect resistant strains in clinical settings. In our study, we employed a range of powerful machine learning tools to predict antibiotic resistance from whole genome sequencing data for E. coli. We used the presence or absence of genes, population structure and isolation year of isolates as predictors, and could attain average precision of 0.92 and recall of 0.83, without prior knowledge about the causal mechanisms. These results demonstrate the potential application of machine learning methods as a diagnostic tool in healthcare settings.

Escherichia coli is a clinically important bacterial species implicated in human- and livestock-associated infections worldwide. The bacterium is known to reside in the guts of humans, livestock, and wild animals.

The rapid horizontal transmission of antibiotic resistance genes on conjugative plasmids between bacterial host cells is a major cause of the accelerating antibiotic resistance crisis. There are currently no experimental platforms for fast and cost-efficient screening of genetic effects on antibiotic resistance transmission by conjugation, which prevents understanding and targeting conjugation. We introduce a novel experimental framework to screen for conjugation-based horizontal transmission of antibiotic resistance between >60,000 pairs of cell populations in parallel. Plasmid-carrying donor strains are constructed in high-throughput. We then mix the resistance plasmid-carrying donors with recipients in a design where only transconjugants can reproduce, measure growth in dense intervals, and extract transmission times as the growth lag. As proof-of-principle, we exhaustively explore chromosomal genes controlling F-plasmid donation within Escherichia coli populations, by screening the Keio deletion collection in high replication. We recover all seven known chromosomal gene mutants affecting conjugation as donors and identify many novel mutants, all of which diminish antibiotic resistance transmission. We validate nine of the novel genes’ effects in liquid mating assays and complement one of the novel genes’ effect on conjugation (rseA). The new framework holds great potential for exhaustive disclosing of candidate targets for helper drugs that delay resistance development in patients and societies and improve the longevity of current and future antibiotics. Further, the platform can easily be adapted to explore interspecies conjugation, plasmid-borne factors, and experimental evolution and be used for rapid construction of strains. IMPORTANCE The rapid transmission of antibiotic resistance genes on conjugative plasmids between bacterial host cells is a major cause of the accelerating antibiotic resistance crisis. There are currently no experimental platforms for fast and cost-efficient screening of genetic effects on antibiotic resistance transmission by conjugation, which prevents understanding and targeting conjugation. We introduce a novel experimental framework to screen for conjugation-based horizontal transmission of antibiotic resistance between >60,000 pairs of cell populations in parallel. As proof-of-principle, we exhaustively explore chromosomal genes controlling F-plasmid donation within E. coli populations. We recover all previously known and many novel chromosomal gene mutants that affect conjugation efficiency. The new framework holds great potential for rapid screening of compounds that decrease transmission. Further, the platform can easily be adapted to explore interspecies conjugation, plasmid-borne factors, and experimental evolution and be used for rapid construction of strains.

Conjugation, the process by which conjugative plasmids are transferred between bacteria, is regarded as a major contributor to the spread of antibiotic resistance, in both environmental and clinical settings. Heavy metals are known to co-select for antibiotic resistance, but the impact of the presence of these metals on conjugation itself is not clear. Here, we systematically investigate the impact that five heavy metals (arsenic, cadmium, copper, manganese, and zinc) have on the transfer of an IncF conjugative plasmid in Escherichia coli. Our results show that two of the metals, cadmium and manganese, have no significant impact, while arsenic and zinc both reduce conjugation efficiency by approximately 2-fold. Copper showed the largest impact, with an almost 100-fold decrease in conjugation efficiency. This was not mediated by any change in transcription from the major Py promoter responsible for transcription of the conjugation machinery genes. Further, we show that in order to have this severe impact on the transfer of the plasmid, copper sulfate needs to be present during the mating process, and we suggest explanations for this.

Predicting bacterial growth from genome sequences is important for a rapid characterization of strains in clinical diagnostics and to disclose candidate novel targets for anti-infective drugs. Previous studies have dissected the relationship between bacterial growth and genotype in mutant libraries for laboratory strains, yet no study so far has examined the predictive power of genome sequence in natural strains.