CORE A ABSTRACT Curation and Statistical Analysis Core Core Director: Kathy Panageas (MSK) The Curation and Statistical Analysis Core will be co-led by Drs. Katherine Panageas and Deb Schrag, who have a long-standing role in the AACR GENIE PRISSMM data curation efforts to date. The PRISSMM framework includes data collection with respect to: Pathology; Radiology; Imaging; Signs and Symptoms; tumor Markers; and Medical oncologist assessments. The phenomic, therapeutic, and oncologic outcome data are curated from the electronic health record (EHR) and other institutional data sources according to the patient- and cancer-centric PRISSMM framework for determination of real-world cancer outcomes. PRISSMM utilizes a cancer-agnostic curation model that standardizes key components of outcome ascertainment in oncology based on existing data standards. Specifically, all pathology reports, all radiology reports for imaging studies other than plain x-rays and ultrasounds, and one medical oncologist note per month are curated. Aim 1 of the Core will be to develop and maintain data standards to support consistent and transparent abstraction of clinical data from electronic health records. Curation initiatives require assessment of different components of data quality throughout the curation process. Multi-pronged QA processes will be implemented and evaluated as part of the Core to address feasibility, accuracy, and reproducibility. In order to assess whether real world progression-free survival (PFS) determined using the PRISSMM data model (based on imaging, PFS-I or based on medical oncologist assessment PFS-M) approximates PFS based on RECIST, the gold standard for defining disease progression in the clinical trial setting, radiologist measurements will be performed and interpretations will be compiled with PRISSMM outcomes. The statistical aspects of the design and analysis of the clinico-genomic analyses in our proposal share many common elements as they relate to analysis of complex genomic data, nuances of PRISSMM data model and patient populations. Aim 2 of the Core will be to provide methodologic and biostatistics support and collaboration for individual observational data analysis. Dedicated shared resources for expert biostatistics and genomics analyses for our P01 program will ensure the necessary rigor and reproducibility critical to our proposed studies.