Machine learning could boost antimicrobial resistance on farms

New research has shown that big data and machine learning could be used to boost antimicrobial resistance (AMR) surveillance on farms, improving outcomes for livestock.

Adobe Stock

Conducted by Nottingham University, the study took place over two and a half years, analysing microbiomes from chickens, carcasses and farm environments. The resulting network of correlations between livestock, environments, microbial communities and antimicrobial resistance suggested multiple possibilities for improving AMR surveillance in livestock production. The work is published in Nature Food.

“The spread of antimicrobial resistant microorganisms and AMR at the human-animal-environment level and food interface is a major global concern,” said Dr Tania Dottorini, Professor of Bioinformatics at Nottingham, and leader of the study. “The transmission of AMR can take place through different routes and pathways, and the food chain, either indirectly via food consumption or directly through contaminated food-animal handling and manure or faecal contamination is a relevant one.

“We have demonstrated how methodologies can be developed that can associate a wide array of microbial species and genes with observable AMR, and further assessed how those are associated with the environmental variables of temperature and humidity. Next, we must consider all relevant and interconnected AMR datasets in a 360° approach, which will deepen our understanding and control of AMR spread.”

The research used a data-mining approach based on machine learning in ten large-scale chicken farms and four connected abattoirs from three provinces in China – one of the largest consumers of antimicrobials. The use of antimicrobials used to prevent and treat infections in livestock production on farms is associated with the rise of AMR infections. AMR is one of the top 10 global public health threats facing humanity according to the World Health Organization, threatening the effective prevention and treatment of an ever-increasing range of infections caused by bacteria, parasites, viruses, and fungi.

“This is an exciting moment,” Dr Dottrorini continued. “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 across animals, environment, humans, and food. This will be groundbreaking.”