Probabilistic graphs to model Pseudomonas aeruginosa survival mechanism and infer low nutrient water response genes

Pseudomonas aeruginosa is an organism notable for its ubiquity in the ecosystem and its antibiotics resistance. This agent presents a particular medical concern because it can live on hospital surfaces and cause various nosocomial infections. Among mechanically ventilated patients with P. aeruginosa pneumonia, ~50% succumb to their condition. Understanding how it survives is important for the design of preventive and curative measures. Furthermore, identifying the survival mechanism in the absence of nutrients is beneficial because P. aeruginosa and related organisms are capable of bioremediation. We hypothesize that P. aeruginosa is capable of long-term survival due to the presence of particular genes which encode for persistence proteins. In this paper, our primary goal is to identify genes responsible for the bacterium's survival. To achieve this, we devised a Bayesian Machine Learning based methodology to analyze the gene expression response to low nutrient water. This approach permitted to learn and construct from gene expression data, an optimal probabilistic graphical model of the survival mechanism. We then used node force techniques to infer a dozen of genes as top orchestrators of the organism's survival mechanism in low nutrient water.