Inferring metabolic networks using the Bayesian adaptive graphical lasso with informative priors

Area of Science:

• Neuroscience
• Systems Biology
• Computational Biology

Background:

• Metabolic processes are crucial for cell function and survival.
• Activated microglia play a key role in neuroinflammation and neurological diseases.
• Understanding microglia metabolism is vital for disease research.

Purpose of the Study:

• To infer a metabolic network in activated microglia using quantified metabolite data.
• To develop and apply a Bayesian adaptive graphical lasso with informative priors for network inference.
• To improve the reliability of network inference, especially with small sample sizes.

Main Methods:

• Application of the Bayesian adaptive graphical lasso with informative priors.
• Utilizing double exponential priors on precision matrix off-diagonal entries for sparsity.
• Formulating tailored hyperpriors on shrinkage parameters to incorporate known biological relationships.
• Leveraging a reference network to guide hyperprior parameter selection.

Main Results:

• Successfully inferred a metabolic network for activated microglia.
• The network incorporates known biological relationships between metabolites.
• Identified novel associations of potential interest for future research.
• Demonstrated improved network inference reliability in small sample size scenarios.

Conclusions:

• The developed Bayesian approach effectively infers metabolic networks in activated microglia.
• This method enhances the understanding of microglia's role in neuroinflammation.
• The inferred network provides valuable insights for neurological disease research.