Machine learning to design antimicrobial combination therapies: promises and pitfalls

Transfer learning predicts species-specific drug interactions in emerging pathogens

Machine learning (ML) algorithms are necessary to efficiently identify potent drug combinations within a large candidate space to combat drug resistance. However, existing ML approaches cannot be applied to emerging and under-studied pathogens with limited training data. To address this, we developed a transfer learning and crowdsourcing framework (TACTIC) to train ML models on data from multiple bacteria. TACTIC was built using 2,965 drug interactions from 12 bacterial strains and outperformed traditional ML models in predicting drug interaction outcomes for species that lack training data. Top TACTIC model features revealed genetic and metabolic factors that influence cross-species and species-specific drug interaction outcomes. Upon analyzing ~600,000 predicted drug interactions across 9 metabolic environments and 18 bacterial strains, we identified a small set of drug interactions that are selectively synergistic against Gram-negative (e.g., A. baumannii) and non-tuberculous mycobacteria (NTM) pathogens. We experimentally validated synergistic drug combinations containing clarithromycin, ampicillin, and mecillinam against M. abscessus, an emerging pathogen with growing levels of antibiotic resistance. Lastly, we leveraged TACTIC to propose selectively synergistic drug combinations to treat bacterial eye infections (endophthalmitis).

Isobolographic analysis of the ciprofloxacin-gentamicin combination against beta-lactamase-producing Staphylococcus aureus

BACKGROUND

Bacterial multi-resistance is a serious global problem that continues to worsen over time due to multiple factors. Among these factors, it is important to highlight the clinical misuse of antibiotics and the mechanisms that microorganisms have developed to protect themselves from these drugs. In this sense, Staphylococcus aureus (S. aureus) is a pathogen that has found a way to resist many of the drugs currently in use, so infections by this bacterium represent a serious clinical problem.

OBJECTIVES

The purpose of this study was to determine the type of interaction between ciprofloxacin and gentamicin against beta-lactamase-producing S. aureus using isobolographic analysis.

METHODS

Ciprofloxacin (0.5-0.05 mg/mL) and gentamicin (10-1 mg/mL) were used to make concentration-dependent curves for each individual drug. Thereafter, the 50 inhibitory concentration (IC50 ) of each drug was obtained, and different proportions of the ciprofloxacin-gentamicin combination-0.5:0.5, 0.8:0.2, 0.2:0.8, 0.9:0.1, 0.1:0.9, 0.95:0.05, and 0.05:0.95-were evaluated. The isobolographic analysis and the interaction index were used to analyze the data.

RESULTS

The isobolographic evaluation of the combination showed that the ratios 0.5:0.5, 0.8:0.2, 0.2:0.8, and 0.9:0.1 produced a synergistic anti-staphylococcal effect, and the 0.95:0.05 ratio induced an additive antibacterial effect. Finally, the 0.1:0.9 and 0.05:0.95 ratios of the combination presented antagonistic effects against S. aureus. On the other hand, the interaction index showed similar results to the isobolographic analysis.

CONCLUSION

The isobolographic results of this in vitro assay show that the ciprofloxacin-gentamicin combination induces synergistic, additive, and antagonistic antimicrobial effects against S. aureus.

New antifungal strategies: drug combination and co-delivery

The growing occurrence of invasive fungal infections and the mounting rates of drug resistance constitute a significant menace to human health. Antifungal drug combinations have garnered substantial interest for their potential to improve therapeutic efficacy, reduce drug doses, reverse, or ameliorate drug resistance. A thorough understanding of the molecular mechanisms underlying antifungal drug resistance and drug combination is key to developing new drug combinations. Here we discuss the mechanisms of antifungal drug resistance and elucidate how to discover potent drug combinations to surmount resistance. We also examine the challenges encountered in developing such combinations and discuss prospects, including advanced drug delivery strategies.

Innovative Strategies to Overcome Antimicrobial Resistance and Tolerance

Antimicrobial resistance and tolerance are natural phenomena that arose due to evolutionary adaptation of microorganisms against various xenobiotic agents. These adaptation mechanisms make the current treatment options challenging as it is increasingly difficult to treat a broad range of infections, associated biofilm formation, intracellular and host adapted microbes, as well as persister cells and microbes in protected niches. Therefore, novel strategies are needed to identify the most promising drug targets to overcome the existing hurdles in the treatment of infectious diseases. Furthermore, discovery of novel drug candidates is also much needed, as few novel antimicrobial drugs have been introduced in the last two decades. In this review, we focus on the strategies that may help in the development of innovative small molecules which can interfere with microbial resistance mechanisms. We also highlight the recent advances in optimization of growth media which mimic host conditions and genome scale molecular analyses of microbial response against antimicrobial agents. Furthermore, we discuss the identification of antibiofilm molecules and their mechanisms of action in the light of the distinct physiology and metabolism of biofilm cells. This review thus provides the most recent advances in host mimicking growth media for effective drug discovery and development of antimicrobial and antibiofilm agents.

A flux-based machine learning model to simulate the impact of pathogen metabolic heterogeneity on drug interactions

Drug combinations are a promising strategy to counter antibiotic resistance. However, current experimental and computational approaches do not account for the entire complexity involved in combination therapy design, such as the effect of pathogen metabolic heterogeneity, changes in the growth environment, drug treatment order, and time interval. To address these limitations, we present a comprehensive approach that uses genome-scale metabolic modeling and machine learning to guide combination therapy design. Our mechanistic approach (a) accommodates diverse data types, (b) accounts for time- and order-specific interactions, and (c) accurately predicts drug interactions in various growth conditions and their robustness to pathogen metabolic heterogeneity. Our approach achieved high accuracy (area under the receiver operating curve (AUROC) = 0.83 for synergy, AUROC = 0.98 for antagonism) in predicting drug interactions for Escherichia coli cultured in 57 metabolic conditions based on experimental validation. The entropy in bacterial metabolic response was predictive of combination therapy outcomes across time scales and growth conditions. Simulation of metabolic heterogeneity using population FBA identified two subpopulations of E. coli cells defined by the levels of three proteins (eno, fadB, and fabD) in glycolysis and lipid metabolism that influence cell tolerance to a broad range of antibiotic combinations. Analysis of the vast landscape of condition-specific drug interactions revealed a set of 24 robustly synergistic drug combinations with potential for clinical use.