Credit: University of Nottingham
Key point:
- Big data and machine learning in antimicrobial resistance (AMR) surveillance identified potential interventions and protections.
- The methodology correlated chicken microbiome species and genes with antibiotic resistance.
- The influence of environmental variables on the microbiome emerged as a target for resistance surveillance.
There are more than a few causes for the concerning rise in antibiotic resistance, with the overuse of antibiotics in livestock production being one of them.
Now, a new study in Nature Food has found that using big data and machine learning in antibiotic resistance surveillance can improve interventions and offer protection against resistant pathogens.
For the study, researchers used a data mining approach based on machine learning in 10 large-scale chicken farms and four connected abattoirs from three provinces in China to analyze microbiomes from chickens, carcasses and environments.
With this approach, the researchers developed a network of correlations between livestock, environments, microbial communities and antibiotic resistance. A subset of the chicken gut microbiome, including antibiotic-resistant genes, correlated with antibiotic resistant profiles of E. coli. Additionally, the gut microbiome makeup was correlated with antimicrobial usage and influenced by environmental temperature and humidity.
“We demonstrated how methodologies can be developed that can associate a wide array of microbial species and genes with observable antimicrobial resistance (AMR), and further assessed how those are associated with the environmental variables of temperature and humidity,” said study author Tania Dottorini, a professor at University of Nottingham.
Targeting the associations between environmental variables and genes associated with resistance may help develop novel antibiotic resistance monitoring solutions, especially in low-middle income countries where chickens are often housed in sheds lacking effective climate control systems.
Dottorini and her team want to take their work a step further by using their datasets in a 360° approach combined with AI.
“We are ready to invest in new AI-powered AMR integrated surveillance approaches to identify the drivers and the mechanisms underlying the insurgence and spread of AMR, and of new genetic variants of resistant pathogens,” she said. “This will be groundbreaking.”