Inferring Gene Regulatory Networks from Multiple Datasets.

Area of Science:

• Computational Biology
• Systems Biology
• Bioinformatics

Background:

• Biological networks are complex and dynamic, requiring advanced inference methods.
• Gene and protein expression data provide insights into cellular processes.
• Traditional models struggle with nonlinearities and large-scale network inference.

Purpose of the Study:

• To provide an overview of Gaussian process dynamical systems (GPDS) for biological network inference.
• To highlight the flexibility and scalability of GPDS for diverse biological data.
• To demonstrate applications in systems biology, synthetic biology, and developmental biology.

Main Methods:

• Utilizing Bayesian nonparametric approaches for nonlinear dynamical systems.
• Applying GPDS to learn biological networks from perturbed time series measurements.
• Employing hierarchical models for inference from multiple datasets with network rewiring.

Main Results:

• GPDS can capture the complexity of ODE models and scale to hundreds of genes.
• Hierarchical GPDS methods allow inference from datasets with changing network structures.
• Leveraging known network structures across species for improved inference.

Conclusions:

• GPDS provide a comprehensive and flexible platform for biological network inference.
• These methods are applicable to systems biology, synthetic biology, and spatiotemporal modeling.
• Tutorials are available to facilitate the application of GPDS.