A cloud-based pipeline for analysis of FHIR and long-read data

Area of Science:

• Genomic Medicine
• Bioinformatics
• Health Informatics

Background:

• Genome sequencing costs are decreasing, making integration with electronic health records feasible.
• Merging genomic and clinical data can enhance clinical decision-making.

Purpose of the Study:

• To demonstrate a comprehensive pipeline for processing PacBio sequencing data and Fast Healthcare Interoperability Resources (FHIR) clinical data.
• To enable secure, scalable tertiary analyses of integrated genomic and clinical information.

Main Methods:

• Utilized PacBio HiFi long-read sequencing data for variant calling via Cromwell on Azure.
• Processed and merged FHIR-formatted electronic health records and genomic data.
• Employed cloud-based Jupyter notebooks for secure tertiary analyses.

Main Results:

• Developed a unified pipeline for handling both genomic and FHIR data.
• Successfully performed tertiary analyses including patient data export, patient clustering, and pharmacogenomic studies.
• Demonstrated the integration of long-read sequencing with standardized clinical data formats.

Conclusions:

• The presented pipeline offers a robust solution for integrating diverse healthcare data types.
• Facilitates advanced analyses for personalized medicine and clinical research.
• The approach supports scalable and secure data processing in a cloud environment.