What does this research mean for the field?

Machine learning models, particularly gradient boosting and deep learning architectures, demonstrate high diagnostic performance for predicting Mycobacterium tuberculosis drug resistance from whole-genome sequencing data, although predicting resistance to certain drugs like pyrazinamide remains challenging. Novelty: ClaimNovelty.SYNTHESIS. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

To systematically evaluate machine learning models that use whole-genome sequencing data to predict drug resistance in Mycobacterium tuberculosis against first-line antibiotics.

August 30, 2026BMC Infectious DiseasesOpen Access

Machine learning models for whole genome based prediction of drug resistance in Mycobacterium tuberculosis: a systematic review

View Full Paper

Ask AI

Bookmark

Authors

HBHadish Bekuretsion AAAssefa Tesfay Abraha GGGebremedhin Gebreslassie

Discussion

Member takes

Overview

Systematic review finds machine learning models accurately predict drug resistance in Mycobacterium tuberculosis from genomic data, highlighting their potential for clinical diagnostics.

Key Points

To systematically evaluate machine learning models that use whole-genome sequencing data to predict drug resistance in Mycobacterium tuberculosis against first-line antibiotics.
Analyzed 15 studies covering 20 distinct machine learning models trained on Mycobacterium tuberculosis whole-genome sequencing data.
Evaluated model architectures (including gradient boosting and attention-based neural networks), feature engineering strategies, bioinformatics pipelines, and performance metrics including sensitivity, specificity, and area under the ROC curve (AUC).
Models demonstrated high sensitivity for major first-line drugs, with 13 of 18 models reporting isoniazid sensitivity achieving ≥90% and 16 of 17 models reporting rifampicin sensitivity achieving ≥90%.
The Hierarchical Attention Neural Network with Task Transfer (HANN-TT) achieved an AUC of 97.9% for isoniazid and 99.1% for rifampicin.
Pyrazinamide resistance was the most challenging to predict across models (sensitivity 56%–98%), though the whole-genome XGBoost (WG-XGB) model reached 95% sensitivity and 99% specificity on the BV-BRC dataset.

Cite This Study

Bekuretsion et al. (2026) studied this question.

synapsesocial.com/papers/6a93f0b06c1a8fb52e79d157 https://doi.org/10.1186/s12879-026-14318-y

View Full Paper

Ask AI

Bookmark

August 30, 2026BMC Infectious DiseasesOpen Access

Machine learning models for whole genome based prediction of drug resistance in Mycobacterium tuberculosis: a systematic review

View Full Paper

Ask AI

Bookmark

Authors

HBHadish Bekuretsion AAAssefa Tesfay Abraha GGGebremedhin Gebreslassie

Discussion

Member takes

Overview

Systematic review finds machine learning models accurately predict drug resistance in Mycobacterium tuberculosis from genomic data, highlighting their potential for clinical diagnostics.

Key Points

To systematically evaluate machine learning models that use whole-genome sequencing data to predict drug resistance in Mycobacterium tuberculosis against first-line antibiotics.
Analyzed 15 studies covering 20 distinct machine learning models trained on Mycobacterium tuberculosis whole-genome sequencing data.
Evaluated model architectures (including gradient boosting and attention-based neural networks), feature engineering strategies, bioinformatics pipelines, and performance metrics including sensitivity, specificity, and area under the ROC curve (AUC).
Models demonstrated high sensitivity for major first-line drugs, with 13 of 18 models reporting isoniazid sensitivity achieving ≥90% and 16 of 17 models reporting rifampicin sensitivity achieving ≥90%.
The Hierarchical Attention Neural Network with Task Transfer (HANN-TT) achieved an AUC of 97.9% for isoniazid and 99.1% for rifampicin.
Pyrazinamide resistance was the most challenging to predict across models (sensitivity 56%–98%), though the whole-genome XGBoost (WG-XGB) model reached 95% sensitivity and 99% specificity on the BV-BRC dataset.

Cite This Study

Bekuretsion et al. (2026) studied this question.

synapsesocial.com/papers/6a93f0b06c1a8fb52e79d157 https://doi.org/10.1186/s12879-026-14318-y

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark