Understanding TB latency using computational and dynamic modelling procedures.

The Mycobacterium tuberculosis bacilli's potency to cause persistent latent infection that is unresponsive to the current cocktail of TB drugs is strongly associated with its ability to adapt to changing intracellular environments, and tolerating, evading and subverting host defence mechanisms. We applied a combination of bioinformatics and mathematical modelling methods to enhance the understanding of TB latency dynamics. Analysis of time course microarray gene expression data was carried out and gene profiles for bacilli adaptation and survival in latency, simulated by hypoxia were determined. Reverse network engineering techniques were used to predict gene dependencies and regulatory interactions. Biochemical systems theory was applied to mathematically model the inferred gene regulatory networks. Significant regulatory genes involved in latency were determined by a combination of systems biology procedures and mathematical modelling of the inferred regulatory networks. Analysis of gene clusters of the inferred networks in the stationary and non-replicating phases of the bacilli predicted probable functions of some of the latency genes to be associated with latency genes of known functions. The systems biology approach and mathematical computational deletion experiments predicted key genes in the TB latency/dormancy program that may be possible TB drug targets. However, these gene candidates require experimental testing and validation.