Machine learning and phylogenetic analysis allow for predicting antibiotic resistance in M. tuberculosis - Research article

Aug 5, 2026

Upload by Oliver

Antimicrobial resistance (AMR) threatens global health, making accurate prediction of bacterial resistance patterns essential for effective treatment and control. Machine learning can analyze large-scale AMR datasets, but it often neglects evolutionary relationships among strains, which may limit detection of resistance-linked features. This study integrates phylogeny information into feature correlation scoring, enabling improved marker discovery and higher predictive performance. Results are validated on Mycobacterium tuberculosis AMR data from PATRIC across six antibiotics.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages12

Reading Time30 min

AccessFree to download

Abstract

Background

Results

Conclusions

Introduction

Public health impact of tuberculosis

Genetic markers and drug mechanisms

Whole-genome sequencing and phylogeny-based methods

Yurtseven et al. BMC Microbiology (2023) 23:404 [https://doi.org/10.1186/s12866-023-03147-7](https://doi.org/10.1186/s12866-023-03147-7) BMC Microbiology RESEARCH Open Access Machine learning and phylogenetic analysis allow for predicting antibiotic resistance in M. tuberculosis Alper Yurtseven1,2*, Sofia Buyanova3, Amay Ajaykumar Agrawal1,2, Olga O. Bochkareva3,4 and Olga V. Kalinina 1,2,5 Abstract Background Antimicrobial resistance (AMR) poses a significant global health threat, and an accurate prediction of bacterial resistance patterns is critical for effective treatment and control strategies. In recent years, machine learning (ML) approaches have emerged as powerful tools for analyzing large-scale bacterial AMR data. However, ML methods often ignore evolutionary relationships among bacterial strains, which can greatly impact performance of the ML methods, especially if resistance-associated features are attempted to be detected. Genome-wide association studies (GWAS) methods like linear mixed models accounts for the evolutionary relationships in bacteria, but they uncover only highly significant variants which have already been reported in literature. Results In this work, we introduce a novel phylogeny-related parallelism score (PRPS), which measures whether a certain feature is correlated with the population structure of a set of samples. We demonstrate that PRPS can be used, in combination with SVM-and random forest-based models, to reduce the number of features in the analysis, while simultaneously increasing models’ performance. We applied our pipeline to publicly available AMR data from PATRIC database for Mycobacterium tuberculosis against six common antibiotics. Conclusions Using our pipeline, we re-discovered known resistance-associated mutations as well as new candidate mutations which can be related to resistance and not previously reported in the literature. We demonstrated that taking into account phylogenetic relationships not only improves the model performance, but also yields more biologically relevant predicted most contributing resistance markers. Keywords Machine learning, Phylogeny, Antimicrobial resistance, Tuberculosis *Correspondence: AlperYurtseven [alper.yurtseven@helmholtz-hips.de](alper.yurtseven@helmholtz-hips.de) 1 Department of Drug Bioinformatics, Helmholtz Institute for Pharmaceutical Research Saarland (HIPS), Helmholtz Centre for Infection Research (HZI), Campus E8 . 1, Saarbrücken 66123, Saarland, Germany 2 Graduate School of Computer Science, Saarland University, Saarbrücken 66123, Saarland, Germany 3 Institute of Science and Technology Austria (ISTA), Am Campus 1, Klosterneuburg 3400, Austria 4 Centre for Microbiology and Environmental Systems Science, Division of Computational System Biology, University of Vienna, Djerassiplatz 1 A, Wien 1030, Austria 5 Faculty of Medicine, Saarland University, Homburg 66421, Saarland, Germany Introduction Mycobacterium tuberculosis (Mtb), the causative agent of tuberculosis (TB), has been a major threat to public health for many years, and remains such a threat now. According to the World Health Organization (WHO), the estimated number of TB-caused deaths in 2021 alone was 1.6 million [1] . TB continues to pose a significant threat to global public health because of its ability to easily transmit and the occurrence of drug-resistant strains of Mtb. In 2019, WHO reposterd over ten million cases, including up to 4.5% of infections with drug resistant bacteria [1] . Early diagnosis and effective treatment are important steps in controlling the spread © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if

FAQs

What challenge does the study address in predicting antibiotic resistance for M. tuberculosis?

It addresses the tendency of machine learning approaches to ignore evolutionary relationships among bacterial strains, which can reduce performance when detecting resistance-associated features.

What is the PRPS method introduced in the study?

The study introduces a phylogeny-related parallelism score (PRPS) that measures whether a feature correlates with the population structure of a set of samples.

How does incorporating phylogenetic relationships affect the results?

Incorporating phylogenetic relationships improves model performance and yields more biologically relevant predicted most contributing resistance markers.

Aug 5, 2026

Upload by Oliver