Enhancing Precision of Pediatric Cancer Molecular Targets by Aggregating CCDI Genomic Data to Pediatric Cancer Knowledgebase

点击翻译按钮获取中文摘要

ABSTRACT Knowledge about molecular targets, including those affected by genetic abnormality, has become increasingly critical for childhood cancer precision oncology. To facilitate exploration of genetic abnormality in pediatric cancer, we developed PeCan (Pediatric Cancer) Knowledgebase in 2015 to establish an open access portal for dynamic visualization of curated somatic variants from published literature contributed by St. Jude and other institutions. PeCanV2, recently released at AACR 2023, now enables navigation of genetic abnormality, including driver mutations and mutational signatures, in the context of ontology of ~300 molecular subtypes of pediatric cancer. By leveraging the PeCanV2 infrastructure, we propose to enhance the precision of FDA- approved Pediatric Molecular Target List (PMTL), characterized by the Childhood Cancer Data Initiative (CCDI), to support clinical applications. This will be achieved by 1) developing an API to generate summary statistics of genetic abnormality at cancer subtype level by aggregating pediatric cancer cohorts profiled by multiple initiatives and harmonized by PeCan (Aim 1); 2) expanding the knowledgebase of pediatric cancer genetic abnormality by integrating three CCDI data sets (i.e. CCDI MSKCC Clinical WGS, CCDI Molecular Characterization and CCDI Metastatic Neuroblastoma), with ~9,000 pediatric cancer samples already harmonized and curated by the St. Jude PeCan portal (Aim 2); and 3) performing mutational signature analysis at cancer subtype level for CCDI samples profiled by whole-genome sequencing using the state-of-art pipelines that we developed (Aim 3). Our team is comprised of researchers and software engineers who have an established track record of leading the world’s largest pediatric cancer genomic studies and data sharing effort. The completion of the proposed work will generate new insights into the molecular drivers of rare or ultra-rare subtypes of pediatric cancer, while the intersection of the PMTL at molecular subtype level will greatly enhance its clinical application. Our proposal thus addresses the key questions put forth by the RFA and is highly relevant to the CCDI Supplement with regards to “data aggregation, linkage, integration, analysis and visualization”, “ultra-rare tumor analysis” and “validation of the FDA Relevant Molecular Target List”.