Children's rare disease cohorts: an integrative research and clinical genomics initiative

Area of Science:

• Genomics
• Bioinformatics
• Rare Diseases

Background:

• Genomic data is often fragmented across different research protocols and clinical settings.
• Harmonizing genomic data, especially for pediatric rare diseases, presents significant challenges for research institutions.
• Integrating diverse data types and stakeholder needs is crucial for effective genomics workflows.

Purpose of the Study:

• To present the outcomes of the Children's Rare Disease Cohorts (CRDC) initiative's first phase at Boston Children's Hospital.
• To establish a unified genomics learning system (GLS) for pediatric rare disease research and clinical practice.
• To create a broadly sharable, harmonized database of genomic and phenotypic data.

Main Methods:

• Developed a genomics learning system (GLS) integrating exome and genome sequencing data.
• Created a relational database for centralized storage and analysis of 8516 exomes and 112 genomes.
• Utilized manual annotation and natural language processing (NLP) for phenotype extraction from clinical records.
• Implemented machine learning algorithms for automated variant classification and prioritization.

Main Results:

• Successfully created a sharable database of 2441 exomes from 15 pediatric rare disease cohorts.
• The GLS integrates genomic, phenotypic, and clinical data, supporting 8516 exomes and 112 genomes.
• Automated variant classification and NLP-based phenotype extraction are integrated within the GLS.
• The platform demonstrates extensibility for future data and analytical system integration.

Conclusions:

• The Children's Rare Disease Cohorts initiative successfully established a harmonized genomics learning system.
• This unified approach facilitates centralized data storage, analysis, and sharing for pediatric rare diseases.
• The GLS provides a robust and extensible platform for advancing genomic research and clinical applications.