Poster #67: Manuel Torres Diz MS, PhD

Arcus Omics: An Enterprise-Scale Platform for Genomic Data and Translational Research at CHOP

Manuel Torres-Diz, MS, PhD1; Khushbu Patel, MS1; Justin Blair, MS1; Arnaud Harmange, MS1; Nicole Feldman, MSIS1; Surendra Buddi, MHI1; Matt Owens, MCS1; Michael Xie, PhD1; Shiva Ganesan, MS1,2,3 ; Yuankun Zhu, BS4; Allison Heath, PhD4; Alexander Gonzalez, PhD1; Scott Haag, PhD1; Spandana Makeneni, PhD1; Heather Griffis, PhD1; Ingo Helbig, MD1,2,3,5 1Dept. of BioMed. and Hlth. Informatics (DBHi), Children’s Hosp. of Philadelphia, 2Div. of Neurology, Children's Hosp. of Philadelphia , 3The Epilepsy NeuroGeneticsInitiative (ENGIN), Children's Hosp. of Philadelphia, 4Center for Data-Driven Discovery in Biomedicine (D3b), Children's Hosp. of Philadelphia, Philadelphia, PA, USA, 5Dept. of Neurology, Univ. of Pennsylvania Perelman Sch. of Med.

The Arcus Omics initiative at Children’s Hospital of Philadelphia (CHOP) is building a robust, enterprise-scale platform that transforms genomic data from clinical and research programs into standardized, discoverable, and reusable research assets. Through the Arcus Omics Data Service, investigators gain secure access to harmonized genomic and clinical datasets supported by rigorous standards for data quality, governance, and provenance. This abstract describes the design, standardization, and implementation of the Arcus Omics Data Service. The Arcus Omics data pipeline manages genomic data throughout its lifecycle. Data generated by clinical and research programs are ingested into the platform where identifiers are standardized, provenance is recorded, processing status is tracked, and source files are securely archived in Amazon Web Services (AWS). Data are then processed through Illumina Connected Analytics (ICA) using DRAGEN pipelines, establishing a consistent analytic platform across CHOP’s genomic ecosystem. Processed outputs are archived and cataloged, while genomic variants and VEP annotations are made queryable in Google BigQuery. Curated datasets can then be provisioned to researchers through Arcus Labs and integrated with clinical data from Epic, enabling efficient genotype-to-phenotype analyses. To date, Arcus Omics has harmonized, processed, and cataloged more than 20,000 genomes and exomes from programs including Undiagnosed Disease Program, Birth Defects Biorepository, Division of Genomic Diagnostics and Epilepsy Genetics Research Project. High-quality SNVs and indels are searchable through Arcus variant browser enabling investigators to rapidly identify variants using flexible, user-specific filtering criteria. Additional workflows support more than 20,000 SNP-array datasets through imputation, further expanding the platforms’ research impact. The Arcus Omics Data Service provides a fully traceable pathway from data generation to scientific discovery, transforming fragmented genomic assets into a unified, scalable, and research-ready eco-system. Combined with the depth and diversity of CHOP’s pediatric patient population, these harmonized genomic and clinical data assets create a unique resource for genotype-phenotype research, rare disease discovery, and precision medicine, while accelerating data-driven and reproducible research across CHOP.