Related Experiment Video
Updated: Jun 14, 2026
07:35
Demonstrating a Multi-drug Resistant Mycobacterium tuberculosis Amplification Microarray
Published on: April 25, 2014
An integrated machine learning approach for predicting DosR-regulated genes in Mycobacterium tuberculosis
Yi Zhang1, Kim A Hatch, Joanna Bacon
1School of Crystallography, Birkbeck College, University of London, Malet Street, London, WC1E 7HX, UK.
BMC Systems Biology
|April 2, 2010
Summary
This study introduces a novel Gaussian process method to infer transcription factor activity (TFA) dynamics from gene expression data. The approach accurately predicts novel target genes regulated by DosR in Mycobacterium tuberculosis, improving our understanding of stress responses.
Area of Science:
- Systems Biology
- Computational Biology
- Genomics
Background:
- DosR regulates Mycobacterium tuberculosis response to stress, like low oxygen.
- mRNA expression doesn't always reflect transcription factor activity (TFA).
- Inferring dynamic TFA is crucial for predicting gene regulation.
Purpose of the Study:
- To develop a method for inferring dynamic transcription factor activity (TFA).
- To model both linear and nonlinear gene responses.
- To predict novel genes regulated by transcription factors.
Main Methods:
- Utilized Gaussian processes (GP) to model dynamic hidden TFAs.
- Applied the GP model to time course gene expression data.
- Integrated promoter sequence information with a logistic regression model.
Main Results:
- Reconstructed the TFA of p53 more accurately than previous methods.
- Successfully estimated DosR TFA in Mycobacterium tuberculosis.
- Predicted ten novel DosR-regulated genes, improving prediction accuracy by integrating promoter data.
Conclusions:
- Gaussian processes effectively reconstruct nonlinear hidden TFA dynamics.
- Chemostat cultures are suitable for studying bacterial responses.
- GP model parameters aid in identifying gene regulation and can be combined with other data for enhanced prediction.