Geography-Aware Optimization of Antimicrobial Combinations Antibiotics are currently the most potent way of fighting bacterial infections but their efficacy is largely undermined by resistance. With the stunted portfolio of new antibiotics, antimicrobial combinations offer an innovative strategy to re-sensitize multidrug-resistant bacterial species to antibiotics they were previously resistant to. However, it is still unclear what antibiotics to combine and how to combine them for synergistic results in clinics. AMR surveillance data is an important but under-utilized tool for prioritizing drug combinations in clinical settings. When two antibiotics are co-administered, their interactions can take one of three directions; it can either be synergistic (positive), antagonistic (negative), or null (neutral). This direction is usually a function of the drug’s mechanism of action, effective concentration, and nature of the pathogen involved. In terms of frequency, antibiotic interactions are neutral (up to 80%), followed by antagonisms (15%), and then synergy (5%). Under prolonged usage, both antagonism and null (95%) interactions have been shown to train ordinary bugs to become superbugs. Therefore, care has to be taken when co-administering drugs in order not to further spread resistance through antibiotic combinations. In this project, we will combine insights from surveillance, drug combination data, and clinical intelligence to predict the best possible combination of antibiotics to fight a given bacterial infection. Specifically, AMR surveillance data containing the resistance phenotypes exhibited by a pathogen will be combined with large antibiotic combination data such as COM-BACT and ARB to generate a graph for finding optimal antibiotic combinations for a given bacterial resistance profile within a defined geographical space. Specifically, we would use Pfizer’s ATLAS datasets which cover a wide geographic and antibiotic space thus making them well-suited for building a global system for personalized management of infections through drug combinations. The success of this project will revolutionize AMR usage and facilitate the rapid reversal of AMR trends by re-sensitizing pathogens to drugs they were previously resistant to.
Center for Global Research Data