Scalable Bayesian Estimation of Multicondition Omics Networks via Clustering and Iterative Merging.

Area of Science:

• Computational Biology
• Systems Biology
• Bioinformatics

Background:

• Inferring condition-specific molecular interaction networks from high-dimensional omics data is crucial for understanding biological mechanisms.
• Challenges include high dimensionality, noise, and accurate cross-condition network comparison.

Purpose of the Study:

• To propose a scalable and flexible Bayesian framework, clustering-focused iterative (CFI) estimation, for joint inference of Gaussian graphical models in multicondition omics data.
• To improve computational efficiency and accuracy in network inference compared to traditional methods.

Main Methods:

• CFI utilizes hierarchical clustering of pooled data to identify consistent subnetwork structures.
• It employs parallel Bayesian estimation within clusters for each condition.
• An iterative merging step recovers intercluster dependencies without restrictive assumptions.

Main Results:

• CFI demonstrated substantial computational gains, with up to 64% average reduction in runtime.
• The framework maintains or improves accuracy compared to traditional approaches for large networks.
• Application to a SARS-CoV-2 proteomics dataset revealed condition-specific network rewiring and identified key biological pathways.

Conclusions:

• CFI is a scalable and accurate method for inferring condition-specific molecular networks from omics data.
• The framework effectively uncovers network rewiring and provides interpretable biological insights.
• CFI advances the analysis of complex biological systems and host-pathogen interactions.