The Monarch Initiative: Linking diseases to model organism resources

点击翻译按钮获取中文摘要

ABSTRACT Biomedical researchers need to identify novel disease genes and understand disease mechanisms; clinicians need to diagnose diseases and optimize treatments. An improved understanding of the genetic basis of disease helps achieve both goals. The Monarch Initiative makes this possible by integrating the fragmented data landscape into the most comprehensive open collection of genotype-phenotype data in the world. Our Knowledge Graph (KG) links together clinical, biomedical, and basic science research data spanning multiple organisms, and supports reasoning across a wide range of organisms, body systems, and diseases. Monarch has achieved demonstrable clinical and translational success using model organism data to perform rare disease diagnosis and gene-to-disease discovery, and our resources have become global standards. In Phase II, we integrated the Human Phenotype Ontology (HPO) into the UMLS; enhanced our variant prioritization algorithms and Exomiser tool, which was applied to 30,000 patients in the National Health Service (UK); developed the Biolink Model/API; released new ontologies (Mondo unified disease ontology, Unified Phenotype Ontology (uPheno), and the Environmental Conditions and Treatments Ontology (ECTO)), raised the number of harmonized data sources in our KG to 34; and overhauled our web Portal. We will leverage this foundation to make Monarch more intuitive for a diversity of users and contexts in phase III as follows: Augment the Monarch Portal with new visualizations and tools. Guided by user requirements, iterative user testing, and feedback, we propose to enhance the user experience and Portal functionality, focusing our work on improvements to Navigation, Visualization, and Query. Evaluate, optimize, and enhance algorithms for disease diagnostics, cross-species inference, and gene-disease discovery. We will develop a comprehensive, modular evaluation framework, ‘PhEval,’ that will allow us to monitor the diagnostic yield and performance of cross-species inference as our ontologies and data graphs evolve. This will assist basic science researchers and clinicians to reveal cross-species mechanistic evidence and evaluate potential precision disease modeling strategies. Disseminate computational tools, data, services, and tutorials to a broad translational community. We will expand access to our KG to enable users to process the KG for different domains and use cases, through simplified downloads, APIs, software libraries, R packages, Jupyter notebooks, and Dockerized resources, along with training materials. This will better support bioinformaticians and other researchers in leveraging our KG and phenotype data in their analyses. The Monarch Initiative aims to significantly improve the utilization, accessibility, and value of animal models for disease diagnosis and discovery.