The Gastric Cancer Registry Genome Explorer: A tool for genomic discovery.

Abstract

434 Background: The Gastric Cancer Registry (GCR) collects clinical questionnaire data and biospecimens from gastric cancer (GC) patients and individuals at high risk for GC (through family history of GC or a germline CDH1 mutation). Its purpose is to facilitate research into better detection and treatment strategies for GC. In 2020, the GCR Genome Explorer (GCR-GE), a publicly accessible and interactive database of clinical and genomic data, was launched to meet this goal. Methods: We generated genomic datasets from participants when GC status and archival gastric tumor tissue were available. We processed the tumor tissue (and paired normal tissue when possible) for whole genome, whole exome, and bulk RNA sequencing. The following genomic features were identified: copy number variations, gene mutations, gene expression, microbiome composition, estimation of tumor-infiltrating immune cells, and tumor neoantigens. We populated (1) all genomic alterations identified from GCR sequencing data, and (2) demographic and pathologic data regarding the tumor that were compiled from participant questionnaires and medical records into the GCR-GE. In addition, sequencing files from the Cancer Genome Atlas (TCGA) were similarly analyzed and uploaded as external datasets. Features of the GCR-GE include cross-study comparisons, study summaries, and queries down to the individual gene, neoantigen, and patient level. Results: In total, 243 GC patients donated tumor samples. The new GCR-GE release contains genomic datasets for 214 of these tumors, as well as datasets for 443 gastric tumors and 185 esophageal tumors from TCGA. Data generation was possible thanks to 756 subjects who enrolled in the GCR from 2011-2022. Most subjects were diagnosed with GC only (N=487), but interestingly, some met multiple criteria. For instance, 10 GC patients had a family history; 10 had a germline CDH1 mutation; and 21 patients had GC, a family history, and a germline CDH1 mutation. Efforts to upload additional datasets are ongoing. All data in the GCR-GE is downloadable. Conclusions: The GCR-GE is a comprehensive resource of genomic and clinical data. Given its accessibility, ease of use, and large cohort sizes, the GCR-GE presents a highly valuable tool for accelerating GC research.