Representation of probabilistic scientific knowledge.

Area of Science:

• Biomedical research
• Computational biology
• Bioinformatics

Background:

• Probability theory is crucial for biomedical data analysis and modeling.
• Previously, hypothesis probabilities were inconsistently recorded as experimental metadata.
• A need exists for standardized methods to represent and reason with probabilities in research.

Purpose of the Study:

• To introduce the Human Experimental Life-science Ontology (HELO) for probabilistic reasoning.
• To enable explicit semantic representation and accurate recording of probabilities associated with research statements.
• To facilitate the inference of methods used to generate and update hypotheses.

Main Methods:

• Development of the HELO ontology to link research statements (hypotheses, models, conclusions) with their probabilities.
• Implementation of semantic descriptors for reporting probabilistic research operations.
• Demonstration of HELO's utility through three case studies.

Main Results:

• HELO successfully links research statements to their associated probabilities.
• The ontology supports the explicit representation of probabilities in hypotheses and inference methods.
• Worked examples showcase HELO's application in analyzing sirtuin-lifespan hypotheses, gene function in S. cerevisiae, and drug design (QSAR).

Conclusions:

• HELO provides a robust framework for probabilistic reasoning in biomedical research.
• The ontology enhances the accuracy and semantic richness of recording research probabilities.
• HELO is an open-source tool that promotes transparency and reproducibility in scientific endeavors.