Background

Beijing Children’s Hospital, Capital Medical University (Beijing Children’s Hospital) is a Grade-A tertiary comprehensive pediatric hospital combining medical care, research, teaching, and healthcare. Its Beijing Pediatric Research Institute is a hospital-run basic-to-clinical pediatric research institute focused on China’s strategic needs in child health. It concentrates on genetic interpretation of hereditary diseases and pathogenic mechanisms; diagnosis of pediatric respiratory infections and related disease mechanisms; child nutrition and nutrition-related diseases; diagnosis and pathogenesis of pediatric ENT and head-and-neck diseases; and diagnosis and etiology of primary immunodeficiency. It contributes to clinical diagnosis, treatment, and mechanistic research on major birth defects, severe respiratory diseases, embryonal tumors, primary immunodeficiency, and pediatric nutrition-related diseases.

Customer Requirements

For medical genetics and genetic diagnosis, the Beijing Pediatric Research Institute needed a precision-medicine platform for hereditary-disease gene interpretation and pathogenic-mechanism research. The platform must meet the many-core, large-memory computing needs of advanced omics assays such as genome sequencing, transcriptome sequencing, proteomics, and metabolomics. Using genomic, proteomic, and other omics technologies together with leading medical methods, it should support biomarker analysis, identification, validation, and application in large cohorts and specific disease types, so researchers can pinpoint disease causes and therapeutic targets, classify disease states and processes precisely, and ultimately deliver personalized precision treatment for diseases and individual pediatric patients—raising the effectiveness of diagnosis, treatment, and prevention and forming high-level medical capability.

Solution

Drawing on years of Sugon technical research and experience in precision medicine, the project delivered an end-to-end, in-depth build: A620-G30 compute nodes, S640-G30 backup nodes, and the Gridview cluster management system.

The A620-G30 is designed for modern data-center workloads. With many cores, large memory, and strong I/O expansion, the compute nodes meet the high-load computing needs of a precision-medicine center and help the data center reach peak performance.

The S640-G30 is designed for high-capacity storage. A single storage node can provide high-density capacity together with strong compute and rich I/O expansion, meeting the precision-medicine center’s need for large-scale backup and archiving.

The Gridview cluster management system provides process scheduling, memory management, system communications, and real-time cluster diagnostics, including a powerful, efficient job scheduler, an integrated simulation-and-design platform, and full-stack data-center operations management.

Customer Value

The project gives the customer a high-performance, easy-to-use, easy-to-operate precision-medicine platform that supports medical genetics and genetic-diagnosis research, simplifies operations for researchers, stores and archives rich clinical and omics research data during operations, and supports experimental analysis and exploratory research by medical and scientific staff at the institute and Beijing Children’s Hospital.