A Bayesian framework for data and hypotheses driven fusion of high throughput data: application to mouse...

Area of Science:

• Computational statistics
• Data science

Background:

• Data integration often involves combining diverse datasets, including experimental data not easily summarized.
• Traditional methods may not be suitable for non-random or complex experimental data.
• Incorporating prior knowledge is crucial for robust data analysis.

Purpose of the Study:

• To present a coherent probabilistic framework for integrating diverse data sets.
• To enable modeling of observed data, parameters of interest, and prior knowledge.
• To facilitate data-driven analysis with complex or experimental data.

Main Methods:

• Development of a probabilistic modeling framework.
• Integration of diverse data sources within a unified setup.
• Incorporation of prior knowledge into the probabilistic model.

Main Results:

• A flexible framework for data integration is established.
• The framework accommodates various data types, including experimental data.
• Prior knowledge can be effectively incorporated into the analysis.

Conclusions:

• The proposed probabilistic framework offers a robust solution for integrating diverse datasets.
• This approach enhances the capability for data-driven analysis, especially with complex experimental data.
• The framework provides a unified and coherent method for handling varied data sources.