Ethan Bustad, Edson Petry, Oliver Gu, Braden T Griebel, Tige R Rustad, David R Sherman, Jason H Yang, Shuyi Ma
{"title":"Predicting fitness in <i>Mycobacterium tuberculosis</i> with transcriptional regulatory network-informed interpretable machine learning.","authors":"Ethan Bustad, Edson Petry, Oliver Gu, Braden T Griebel, Tige R Rustad, David R Sherman, Jason H Yang, Shuyi Ma","doi":"10.3389/ftubr.2025.1500899","DOIUrl":null,"url":null,"abstract":"<p><strong>Introduction: </strong><i>Mycobacterium tuberculosis</i> (Mtb) is the causative agent of tuberculosis disease, the greatest source of global mortality by a bacterial pathogen. Mtb adapts and responds to diverse stresses, such as antibiotics, by inducing transcriptional stress response regulatory programs. Understanding how and when mycobacterial regulatory programs are activated could inform novel treatment strategies that hinder stress adaptation and potentiate the efficacy of new and existing drugs. Here, we sought to define and analyze Mtb regulatory programs that modulate bacterial fitness under stress.</p><p><strong>Methods: </strong>We assembled a large Mtb RNA expression compendium and applied this to infer a comprehensive Mtb transcriptional regulatory network and compute condition-specific transcription factor activity (TFA) profiles. Using transcriptomic and functional genomics data, we trained an interpretable machine learning model that predicts Mtb fitness from TFA profiles.</p><p><strong>Results: </strong>We demonstrated that a TFA-based model can predict Mtb growth arrest and growth resumption under hypoxia and reaeration using gene expression data alone. This model also directly elucidates the transcriptional programs driving these growth phenotypes.</p><p><strong>Discussion: </strong>These integrative network modeling and machine learning analyses enable the prediction of mycobacterial fitness across different environmental and genetic contexts with mechanistic detail. We envision these models can inform the future design of prognostic assays and therapeutic interventions that can cripple Mtb growth and survival to cure tuberculosis disease.</p>","PeriodicalId":101354,"journal":{"name":"Frontiers in tuberculosis","volume":"3 ","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12269550/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Frontiers in tuberculosis","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.3389/ftubr.2025.1500899","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/4/2 0:00:00","PubModel":"Epub","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Introduction: Mycobacterium tuberculosis (Mtb) is the causative agent of tuberculosis disease, the greatest source of global mortality by a bacterial pathogen. Mtb adapts and responds to diverse stresses, such as antibiotics, by inducing transcriptional stress response regulatory programs. Understanding how and when mycobacterial regulatory programs are activated could inform novel treatment strategies that hinder stress adaptation and potentiate the efficacy of new and existing drugs. Here, we sought to define and analyze Mtb regulatory programs that modulate bacterial fitness under stress.
Methods: We assembled a large Mtb RNA expression compendium and applied this to infer a comprehensive Mtb transcriptional regulatory network and compute condition-specific transcription factor activity (TFA) profiles. Using transcriptomic and functional genomics data, we trained an interpretable machine learning model that predicts Mtb fitness from TFA profiles.
Results: We demonstrated that a TFA-based model can predict Mtb growth arrest and growth resumption under hypoxia and reaeration using gene expression data alone. This model also directly elucidates the transcriptional programs driving these growth phenotypes.
Discussion: These integrative network modeling and machine learning analyses enable the prediction of mycobacterial fitness across different environmental and genetic contexts with mechanistic detail. We envision these models can inform the future design of prognostic assays and therapeutic interventions that can cripple Mtb growth and survival to cure tuberculosis disease.