Computationally Efficient Bayesian Estimation of Graphical Networks for Omics Data

Area of Science:

• Computational Biology
• Bioinformatics
• Systems Biology

Background:

• Graphical networks model complex biological processes using omics data.
• Bayesian analyses offer advantages over frequentist methods for omics data, including prior knowledge integration.
• Current Bayesian approaches are computationally intensive, limiting their use with large omics datasets.

Purpose of the Study:

• To present BPlane (Bayesian PseudoLikelihood-based Algorithm for Network Estimation), a computationally efficient Bayesian method.
• To extend Bayesian network modeling capabilities for large-scale omics data, such as untargeted proteomics.
• To demonstrate BPlane's computational savings and accuracy compared to existing methods.

Main Methods:

• Developed BPlane, a Bayesian pseudo-likelihood-based algorithm for network estimation.
• Evaluated BPlane's performance using simulations.
• Applied BPlane to a SARS-CoV-2 proteomics dataset.

Main Results:

• BPlane achieves substantial computational savings compared to state-of-the-art Bayesian algorithms.
• BPlane maintains competitive edge detection accuracy.
• Demonstrated computational benefits on a large SARS-CoV-2 proteomics dataset (7000 proteins).

Conclusions:

• BPlane significantly enhances the feasibility of Bayesian network analysis for large omics datasets.
• The method offers a computationally efficient solution for biological network estimation in proteomics and other omics fields.
• BPlane facilitates deeper insights into complex biological systems through scalable Bayesian modeling.