Related Experiment Video
Updated: Jul 10, 2026
07:35
Selecting Multiple Biomarker Subsets with Similarly Effective Binary Classification Performances
Published on: October 11, 2018
stackPredAMR-a stacked random forest approach improves AMR phenotype prediction for multiple species and
Julian Welling1,2, Miriam Balzer1,2, Leah Consten1
1Department of Medicine, University of Duisburg-Essen, Essen, 45147, Germany.
Bioinformatics Advances
|July 9, 2026
Summary
A new machine learning model, stackPredAMR, accurately predicts antimicrobial resistance in key bacterial pathogens using whole genome sequencing data. This rapid method offers a significant advancement over traditional testing for combating antimicrobial resistance.
Area of Science:
- Microbiology
- Bioinformatics
- Machine Learning
Background:
- Antimicrobial resistance (AMR) is a critical global health challenge.
- Current antimicrobial susceptibility testing (AST) methods are slow due to reliance on bacterial cultivation.
- Whole genome sequencing (WGS) with machine learning (ML) presents a faster alternative but faces limitations in scope and data.
Purpose of the Study:
- To develop and validate a novel ML framework, stackPredAMR, for predicting AMR.
- To enhance prediction accuracy by incorporating antimicrobial resistance gene presence and cross-resistance patterns.
- To provide a scalable and extensible solution for rapid AMR detection.
Main Methods:
- Developed stackPredAMR, an ML framework utilizing a stacked architecture with random forest layers.
- Input features include antimicrobial resistance gene presence.
- Model trained and benchmarked on over 2500 WGS datasets linked to phenotypic resistance data for *Escherichia coli*, *Klebsiella pneumoniae*, and *Acinetobacter baumannii*.
Main Results:
- stackPredAMR achieved high performance metrics: median accuracy of 0.94, ROC AUC of 0.97, and F1-score of 0.91.
- The model demonstrated superior performance compared to existing methods.
- Predicted resistance to 18 antimicrobial agents across three bacterial species.
Conclusions:
- stackPredAMR offers a rapid, accurate, and cost-effective method for predicting antimicrobial resistance.
- The framework's design supports future expansion to more species and antimicrobial agents.
- Freely available source code and datasets facilitate adoption and further research.
