Investigating the pervasiveness and distribution of known and unknown antimicrobial resistance genes, coupled with toxic metabolite production, in Qatar: A comprehensive study across various sectors for AI modeling.

Antimicrobial resistance (AMR) will cause at least another 100 million deaths by 2030. Newer antibiotics are limited, and thus, we must find alternative approaches to stop the spread of drug-resistant bacteria. To combat AMR, colistin has been repurposed as a last-resort antibiotic to treat patients with bacterial infections, under controlled conditions to avoid toxicity. The misuse of colistin as a growth promoter in animal husbandry has led to the emergence of new resistance mechanisms. Bacteria can become resistant to colistin by acquiring the gene mcr-1. We hypothesize that selection pressures from the misuse of colistin in various sectors can lead to bacteria with elevated resistance to colistin. We isolated E. coli from wastewater showing escalated resistance to colistin. We expect these bacteria to harbour new genetic factors contributing to the enhanced colistin resistance. Our findings will be used to assemble a multimodal dataset—phenotypes —into a harmonized data lake. We will train AI models to predict AMR-gene prevalence, map hotspots across wastewater networks, and run “what-if” simulations to guide stewardship that curtails spread in remedial and pristine sites in Qatar. In addition, these AI models would be critical for stakeholders, such as MoPH, CDC, and Baladna, to institute policy changes that would save humans and the animal, poultry, fish, and agriculture industries from infectious bacteria.

Hamad Bin Khalifa University (HBKU)