Researchers from the University of the Philippines Diliman are using artificial intelligence (AI) to help predict antibiotic resistance in Escherichia coli (E. coli), a common bacterium that can resist antibiotics and cause health problems. The team tested different AI models to analyze genetic data and lab test results from the National Center for Biotechnology Information (NCBI) database.
The study was conducted by Dr. Pierangeli Vital from the Natural Sciences Research Institute (NSRI), along with Marco Christopher Lopez and Dr. Joseph Ryan Lansangan from the UPD School of Statistics.
“We selected the models based on their strengths in handling biological and imbalanced data,” said Vital. “These models were chosen to compare performance across different learning strategies and to identify which is most suitable for predicting antibiotic resistance.”
The AI models tested include Random Forest (RF), Support Vector Machine (SVM), Adaptive Boosting (AB), and Extreme Gradient Boosting (XGB). RF is known for handling large and complex data. SVM works well when it comes to separating different groups of data, especially in tricky cases. AB and XGB are advanced techniques that combine multiple predictions to improve results, especially when some samples are hard to classify.
The team explained that traditional testing methods for antimicrobial resistance can take a lot of time and effort. These tests may not be practical for large-scale use, especially in places like farms where early detection is important.
The AI models were able to predict resistance to some antibiotics better than others. “Most accurately,” the models predicted resistance to streptomycin and tetracycline, but had difficulty with ciprofloxacin due to fewer resistant samples in the data. Among the models, AB and XGB gave consistently better results even with data that was not balanced.
“We think that this strategy has great potential for real-time monitoring of antimicrobial resistance, particularly in agriculture,” Vital said. “As DNA sequencing becomes faster and cheaper, prediction models such as ours can pick up resistant bacteria early, before they lead to outbreaks. This can facilitate better decision-making in food safety, agriculture, and public health programs.”
The researchers suggest adding more types of samples and data, such as metagenomic data, to improve predictions and better understand how bacteria develop resistance.
Image from the University of the Philippines College of Science
Get the latest before it trends. Follow Back End News on LinkedIn, Facebook, X, YouTube, and TikTok for updates and in-depth coverage across the tech and security landscape.