An ontology for Autism Spectrum Disorder (ASD) to infer ASD phenotypes from Autism Diagnostic Interview-Revised data

Area of Science:

• Computational biology
• Ontology engineering
• Neuroscience

Background:

• Autism Spectrum Disorder (ASD) and related neurodevelopmental disorders (NDD) require robust diagnostic tools.
• Integrating diverse subject data is crucial for advancing research in ASD and NDD.
• Existing ontologies may not fully capture the complexity of ASD phenotypes and diagnostic criteria.

Purpose of the Study:

• To develop an ontology for data integration and reasoning in ASD and NDD research.
• To automatically infer ASD phenotypes and Diagnostic & Statistical Manual of Mental Disorders (DSM) criteria using subject assessment data.
• To establish a foundational step towards a comprehensive knowledge base for ASD and related disorders.

Main Methods:

• Augmented an existing autism ontology with knowledge on diagnostic instruments, ASD phenotypes, and risk factors using OWL and semantic web rules.
• Developed a custom Protégé plugin to manage combinatorial OWL axioms for many-to-many relationships between ADI-R items and DSM diagnostic categories.
• Utilized a reasoner to infer DSM-IV-TR and DSM-5 diagnostic criteria for 2642 subjects based on their Autism Diagnostic Interview-Revised (ADI-R) data.

Main Results:

• Extended the ontology with 443 classes and 632 rules, incorporating phenotypes, synonyms, risk factors, and comorbidity frequencies.
• Achieved high accuracy in inferring diagnoses: true positive and true negative rates of 1 and 0.065 for DSM-IV autistic disorder, and a 0.94 true positive rate for DSM-5 ASD.
• Demonstrated the ontology's capability for accurate automatic inference of disease phenotypes and diagnoses.

Conclusions:

• The developed ontology accurately infers subject disease phenotypes and diagnoses.
• This ontology serves as a valuable knowledge base for ASD research.
• Future expansion to include related NDDs can facilitate automatic inference of commonalities and differences, enhancing understanding of ASD pathophysiology.