Abstract
The American Association for Cancer Research (AACR) Project GENIE (Genomics Evidence Neoplasia Information Exchange) Biopharma Collaborative (BPC) is a multi-phase, pre-competitive collaboration between 10 biopharmaceutical companies and select GENIE-participating academic institutions, focused on detailed clinical annotations of a subset of patients within the GENIE Registry. The cohorts focus on 10 solid tumors, and each integrates demographic, diagnosis, genomic, and treatment data with longitudinal, real-world patient outcomes. Data are collected following a structured framework to ensure interoperability and forward compatibility with other data models. Each cohort undergoes a series of rigorous quality control and assurance protocols which ensures consistency, accuracy, and reliability of the data across multiple institutions before public release of the data. Initial analyses of the BPC data have yielded valuable insights, including the validation of treatment-induced resistance mutations and genomic drivers associated with anatomic sites of metastasis. Additionally, the real-world response endpoints compare favorably to published trial results. Central management and a shared knowledgebase help integrate diverse functional teams in the execution of a complex, multi-institutional data collection effort. Future directions aim to automate significant portions of the clinical annotation process to collect clinical data at scale. These efforts will increase the depth and granularity of the BPC data, as well as expand the overall cohort size and range of cancer types represented.
Key words: AACR Project GENIE, real-world data, precision medicine, clinico-genomic data, longitudinal data, real-world patient outcomes
Highlights
-
•
GENIE BPC is an academic biopharma initiative to provide publicly available high quality real-world clinical-genomic data.
-
•
The project generated 10 cancer cohorts integrating real-world genomic, patient demographic, and treatment data with real-world PFS and OS.
-
•
Internal analyses have validated treatment-induced resistance mutations, genomic drivers of sites of metastasis, and published trial data.
-
•
Each of the cohorts will be made publicly available, with several currently available for community use.
Introduction and rationale
The American Association for Cancer Research (AACR) Project GENIE (Genomics Evidence Neoplasia Information Exchange) is a publicly available, international, pan-cancer registry of real-world clinico-genomic data generated through data sharing amongst a consortium of tertiary academic referral centers.1, 2, 3
The GENIE registry contains clinical-grade, oncologic, next-generation sequencing (NGS) test results along with limited clinical annotation of 19 data elements for nearly 185 000 patients and >214 000 sequenced samples as of the 16.1-public release.4 The size of the registry makes it an invaluable resource for the research community with >1400 citations and nearly 19 000 registered users as of 22 October 2024; however, a deeper level of clinical annotation on each patient is required to advance precision oncology. To this end, the GENIE Consortium embarked on a multiyear collaboration with 9 large biopharmaceutical companies, later expanding to 10, collectively named the Biopharma Collaborative (BPC), to curate extensive clinical data on specific cancer types.5,6
The BPC uses a precompetitive model to reflect the guiding principles of open science, transparency, and inclusion embodied by GENIE. In this model, both biopharmaceutical partners and participating institutions have access to the data during an embargo period before public release.7,8 Decision-making and governance are shared through the BPC subcommittee consisting of pharmaceutical partners, academic institutions, and select project members, who collaboratively oversee project direction while a smaller ‘core team’ is responsible for the daily operations and execution. Participating BPC individuals, companies, and institutions are committed to making the BPC data public, which is increasingly important in the era of precision medicine.
Data curation and processing
The BPC subcommittee selected non-small-cell lung (NSCLC), colorectal (CRC), early-onset breast, prostate, pancreatic, and bladder cancers as the cohorts in phase I (curated 2019-2021) of the BPC. In phase II (curated 2022-2025), the NSCLC and CRC phase I cohorts were updated with additional variables, extended patient follow up, and expanded with new patients, doubling the size of each cohort to >3000 patients (Table 1). Data on patients with renal cell, ovarian, cutaneous melanoma, and esophagogastric cancers will also be curated during phase II. Patients in the GENIE Registry with a confirmed index cancer diagnosis, who were ≥18 years of age at the time of genomic sequencing, which was conducted >2 years before commencement of data collection, were randomly selected for inclusion in each cohort, with the exception of the early-onset breast cancer cohort, who were <56 years old at the time of genomic sequencing [Figure 1A(iv)]. Phase I included patients from the Memorial Sloan Kettering Cancer Center, Dana Farber Cancer Institute, Vanderbilt-Ingram Cancer Center, and the University Health Network Princess Margaret Cancer Centre. Phase II added Providence Cancer Institute; Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins; University of California-San Francisco Helen Diller Family Comprehensive Cancer Center; and Vall d’ Hebron Institute of Oncology from the broader GENIE Consortium. Patients in the GENIE Registry have received at least one clinical-grade NGS test at a Consortium Institution as part of their care journeys and therefore, the registry and cohorts within the registry, may not reflect national or international demographics for any given cancer.
Table 1.
Additional data elements collected
| Cohort (n) | Additional data elements |
|---|---|
| NSCLC (3844) | Smoking status, separate tumor nodules ipsilateral lung, pleural invasion by H&E or elastic stain, radiation treatment, performance status, and PD-L1 |
| CRC (3010) | Circumferential radial margin, perineural invasion, tumor deposits, radiation treatment, performance status, microsatellite instability status, mismatch repair status, CA19-9, and CEA |
| Breast (1126) | ER/PR/HER2 by any method on all specimens, HER2 summary, multigene result, multigene method, Oncotype Dx Risk Level-Invasive, CA15-3, CEA, and CA27.29 |
| Pancreas (1124) | Tumor resectability status, radiation treatment, mismatch repair status, CA19-9, and CEA |
| Prostate (1125) | Medical oncologist assessment of hormone/castrate/androgen-independent status; Gleason patterns and score (clinical and pathologic), Gleason tertiary pattern and score, number of cores positive and examined, radiation treatment, PSA, and testosterone |
| Bladder (750) | Smoking status, tumor invasion of the muscularis propria, performance status, and radiation treatment |
| Renal Cell (1143) Carcinoma | Smoking status, invasion beyond capsule, ipsilateral adrenal gland involvement, major vein involvement, sarcomatoid and rhabdoid features, IMDC risk score, performance status, and radiation treatment |
| Ovarian (1286) | FIGO stage, CA-125 pretreatment interpretation, CA-125 tumor markers, residual tumor volume post cytoreduction, chemotherapy response score, summary of known pathogenic variants, performance status, and radiation treatment |
| Melanoma (1252) | Data dictionary under development |
| Esophagogastric (1250) | Data dictionary under development |
Cancer-specific data elements were defined for each BPC cohort and added to the PRISSMM framework. Selection of these cancer-specific variables relied upon the North American Association of Central Cancer Registries (NAACCR) Site-Specific Data Item (SSDI) Manual, review of relevant literature on oncology data standards, and input from multidisciplinary teams comprised of clinical oncologists, investigators, statisticians, and informatics experts.21 These sources informed the identification of pertinent datapoints specific to each cancer type, ensuring comprehensive coverage of clinically relevant information. Target case numbers (n) may differ slightly from public releases.
BPC, Biopharma Collaborative; CEA, carcinoembryonic antigen; ER, estrogen receptor; FIGO, International Federation of Gynecology and Obstetrics; H&E, hematoxylin–eosin; HER2, human epidermal growth factor 2; IMDC, International Metastatic renal cell carcinoma Database Consortium; PD-L1, programmed death-ligand 1; PR, progesterone receptor; PSA, prostate-specific antigen.
Figure 1.
(A) (i) GENIE participating institutions upload variant calls, structural and copy number data, BED files, workflow files, and 19 clinical data elements to the Synapse platform. Data are processed, harmonized, filtered, and internally reviewed (ii) before being made available through both cBioPortal and Synapse (iii). Patients for Biopharma Collaborative (BPC) cohorts are selected for additional clinical annotation using criteria collaboratively decided by the BPC subcommittee (iv). Patients selected for BPC cohorts are clinically annotated using the PRISSMM framework with additional data elements curated for specific cancer types (v). Once data curation is complete, additional variables are derived and programmatic data quality checks are iteratively reported and resolved by quality assurance managers at each institution (vi). At data lock, the data are uploaded, processed, and associated with existing genomic data files in the registry (i). (B) BPC workflow process. During preproduction, the cancer cohort is selected, and the data dictionary updated (pilot testing) and tested (pressure testing) before moving to production. In production, structured (database icon) and abstracted data (document icon) are entered into REDCap, where integrated data quality rules are applied (i-ii) and a random sample of 20% of production cases are selected for source data verification (iii). Once curation of a cohort is complete, data are de-identified and investigational agents are masked (iv) before upload (v) where data are ingested (vi), integrity checks are carried out, and data tables created (vii) before several sequential rounds of programmatic data quality checks and query resolution are carried out to address complex data quality issues (vii-x). Once data for a cohort are locked, the versioned data release is made available on cBioPortal and Synapse (xi). (C) BPC patient timeline visualization in cBioPortal showing biopsies and resections, diagnoses, treatments, as well as assessments from imaging and medical oncology (red, amber, green, and grey circles) (i). The genomic event tables (ii) show detailed information about mutations, structural variants, and copy number alterations along with the protein change, variant allele frequency, and pathogenicity mapped to the sample. (iii) The tooltip is shown for the OncoKB annotation (blue target) of ALK G1202R, highlighting a known resistance mutation to crizotinib ©de Bruijn et al.16.; Published by the American Association for Cancer Research. (D) Each BPC cohort contains data on therapies given before and after next generation sequencing (NGS) testing. (i) comparison of the genomic alterations of patients with NGS testing on samples collected before and after erlotinib and before and after crizotinib (ii) reveal frequent post-treatment alterations. Lollipop plots and inset boxes highlight acquisition of known mechanisms of resistance, including second-site epidermal growth factor receptor alterations, acquired gene fusions, and MET amplification after erlotinib (i). Acquisition of the analogous gatekeeper mutations ROS1 G2032RR, and ALK G1202R after crizotinib (ii) ©Choudhury et al.17; Published by the American Association for Cancer Research.
BED, browser extensible data; BrCa, breast cancer; CNA, copy number alteration; CRC, colorectal cancer; MAF, mutation annotation file; NSCLC, non-small-cell lung carcinoma; PANC, pancreatic cancer; PrCa, prostate cancer; QA, quality assurance; SDV, source data verification; VCF, variant call file; XRT, radiotherapy.
The BPC datasets consist of patient- and sample-level data sourced from the AACR Project GENIE Registry, supplemented with additional longitudinal clinical data collected from the institutional electronic health records and tumor registries into local REDCap databases by curators utilizing the PRISSMM™ framework.9,10 PRISSMM™ standardizes data collection from Pathology, Radiology, Imaging, Signs and Symptoms, tumor Markers, and Medical oncologist assessments and combines these data with patient demographics and treatments [Figure 1A(v)]. Serum concentrations of standard-of-care protein biomarkers [e.g. prostate specific antigen (PSA)] and pathologically confirmed biomarkers (e.g. programmed death-ligand 1, mismatch repair/microsatellite instability, estrogen receptor/progesterone receptor/human epidermal growth factor 2), as well as cancer-directed radiation therapy data along with other data elements were curated for select cohorts (Table 1). In instances where patients had a history of multiple cancer diagnoses, all pertinent clinical elements were curated for each cancer diagnosis, including associated treatments. Additional variables, including overall and progression-free survival endpoints, along with an accompanying analytic data guide, are subsequently generated by the Statistical Coordinating Center at Memorial Sloan Kettering Cancer Center and included in each data release.11 All project data are de-identified and the name and duration of investigational agents are masked to protect intellectual property [Figure 1B(iv)].
The PRISSMM™ framework uses standardized vocabularies and data structures. End users can transform the BPC datasets to other data models such as the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM). The data tables ‘patient-characteristics’ and ‘ca_directed_drugs’ tables, for example, are equivalent to the OMOP ‘person’ and ‘drug_exposures’ tables, respectively. Data users can similarly transform the BPC datasets into other data models.
Rigorous quality assurance (QA) procedures were instituted and systematically integrated in the creation of each BPC cohort.12 Before data curation, a pilot phase assesses the feasibility of collecting each cancer-specific data element, refining curation directives, and providing additional training for curators before initiating cross-institutional testing. Data quality rules were integrated into the REDCap project to identify instances of missing data or implausible values during curation [Figure 1B(i)]. To ensure data reliability, source data verification was implemented for a randomly selected 20% sample of cases, where QA managers at each participating institution verified the accuracy of curated clinical data against the original sources [Figure 1B(iii)]. Further, a randomly selected 5% sample of cases was curated by two independent curators to evaluate the reproducibility of the curation directives. Post-production QA processes were developed to programmatically evaluate complex data quality issues through multiple rounds of query, review, and resolution until all queries were addressed [Figure 1B(vii-x)].
Data releases, results, and future directions
Data from each cohort are independently released to biopharmaceutical partners and core team members for an initial 6-month period. After this period, investigators at any GENIE participating institution can use the data for an additional 6 months before public release. Requests for data access by consortium members are reviewed by respective institutional steering committee members and are centrally managed by AACR. Consortium members are granted access to a subset of all variables immediately through cBioPortal, whereas requests for complete data from an embargoed dataset require submission and approval of a research proposal overseen by the data use and publications committee.1 Data are accessible via the Synapse platform and a dedicated cBioPortal instance [Figure 1B(xi)]; further, the genieBPC package for R statistical software pulls data from Synapse and can be used to generate an analytic cohort.13, 14, 15 Close collaboration with the cBioPortal development team ensures continued enhancements and support of an expanding array of data types and exploratory analyses [Figure 1C(i)].16
Multidisciplinary academic teams, including disease experts, are assembled to carry out a comprehensive analysis for each cohort. In the BPC NSCLC 2.0-public dataset, 44% of patients had actionable mutations in their tumors, with distinct patterns of co-mutations and genomic signatures of smoking validated.17 Real-world overall survival and progression-free survival were reported for the entire cohort and for patients receiving various treatment regimens, including platinum-based therapies and immune checkpoint inhibitors. These results compared favorably with major clinical trials including PARAMOUNT and KEYNOTE-010.17 Finally, genomic drivers associated with sites of metastasis, and the emergence of treatment-induced resistance mutations were described (Figure 1D).17 Similar analyses in the early-onset breast cancer cohort reveal patterns of gene alterations and overall survival by subtype and age group, as well as genomic drivers of sites of metastasis (unpublished results).
As the first six BPC cohorts are publicly released, we will encourage analyses by the community through AACR journal programs and other mechanisms, including a call for invited presentations at the AACR annual meeting that utilize these data. Several internal analyses are underway including off-label/off-guideline drug use, patterns of clinical trial participation, and an evaluation of the timing of sequencing relative to diagnosis across all six cohorts has been published.18
Conclusion
The BPC is a multiyear, multicenter project with numerous deliverables across 10 cancer types. Collecting and standardizing high-quality, longitudinal clinical annotations of highly salient treatment outcomes across institutions is complex and challenging. We have found that a centralized knowledgebase to ensure continuity of information; coordination of efforts by the various teams; clear and effective communications between and among the functional teams and the project leadership; and bundling of related activities to reduce the number of handoffs between teams has been successful. Future work includes using BPC data to implement various programmatic methodologies to enhance automation and thereby expedite and expand capacity for clinical annotation across the entire GENIE Registry; early single institution experiences suggest that this is feasible.19,20 The team looks forward to providing high-quality, longitudinal data that combine genomic features, treatment exposures, and clinical outcomes data on additional patients to the broader research community.
Funding
AACR Project GENIE receives research funding from Amgen, AstraZeneca, Bayer, Bristol Myers Squibb, Boehringer Ingelheim, Genentech, Janssen, Merck, Novartis, and Pfizer.
Disclosure
AA reports institutional support for AACR Project GENIE BPC from Genentech, Novartis, Amgen, Merck, Bayer, Janssen, Bohringer Ingelheim, Bristol Myers Squibb (BMS), AstraZeneca, and Pfizer. PLB reports institutional support from AACR Project GENIE, BMS, Sanofi, AstraZeneca, Amgen, Pfizer, Genentech/Roche, GlaxoSmithKline, Novartis, Merck, Medicenna, Seagen, Zymeworks, Lilly, Amgen, Bicara Therapeutics, Takeda, and LegoChem; participation in data safety and monitoring board for Roche/Genentech, Lilly, Canadian Myeloma Research Group; leadership roles in AACR Project GENIE and Breast International Group Executive Board; medical writing support from Genentech/Roche, Novartis, Merck, Medicenna, Seagen, Zymeworks, Lilly, Takeda; and uncompensated advisory for BMS, Pfizer, Seagen, Lilly, Amgen, Merck, Zymeworks, Gilead, Janssen. SB received travel and accommodation expenses from AACR Project GENIE. MF reports institutional support for AACR Project GENIE BPC from Genentech, Novartis, Amgen, Merck, Bayer, Janssen, Bohringer Ingelheim, BMS, AstraZeneca, and Pfizer. JNH reports institutional support for AACR Project GENIE BPC from Genentech, Novartis, Amgen, Merck, Bayer, Janssen, Bohringer Ingelheim, BMS, AstraZeneca, and Pfizer. KLK reports institutional support from AACR Project GENIE; and travel and accommodation expenses from AACR Project GENIE. EL reports employed from Dana Farber Cancer Institute; royalty payments for Dana Farber Cancer Institute for PRISSMM. MLLN reports institutional support from AACR Project GENIE and GE Healthcare; travel and accommodation expenses from AACR Project GENIE; and a pending patent with GE Healthcare. CMM reports institutional research support from AACR Project GENIE; travel and accommodation expenses from AACR Project GENIE. CN reports employment from AACR Project GENIE; travel and accommodation expenses from AACR Project GENIE. KSP reports institutional support from AACR Project GENIE; NIH P30 CA008748 grant; travel and accommodation expenses from AACR Project GENIE; and stock interests in 23and Me, Vincerx, Eyepoint, and Kyverna. BP reports institutional support from AACR Project GENIE, Illumina, and Lilly; has provided speaker services for Illumina. RP reports institutional support from AACR Project GENIE; travel and meeting expenses from AACR Project GENIE. GJR reports institutional support from AACR Project GENIE, Mirati, Lilly, Takeda, Merck, Roche, Pfizer, Novartis; has provided speaker services for DAVA Oncology, Imedex, MJH Associates, NCCN, Research to Practice, Triptych Health Partners; participation in data and safety monitoring board for Novartis; and receipt of equipment, materials, drugs, medical writing, gifts or other services from Amgen. HR reports employment by AstraZeneca (not during study period). DS reports institutional research funding from AACR Project GENIE; royalties for development of PRISSMM curation framework; and travel and accommodation expenses from AACR Project GENIE. SMS reports institutional research funding from Amgen, Inc., AstraZeneca UK Limited, Bayer Healthcare Pharmaceuticals, Inc., Boehringer Ingelheim, GMBH, BMS, Genentech, Janssen R&D, LLC now J&J Innovative Medicines, Merck, Sharpe & Dohme, LLC, Novartis and Pfizer, Inc. All other authors have declared no conflicts of interest.
Acknowledgements
The authors acknowledge the American Association for Cancer Research (AACR) and its financial and material support in the development of the AACR Project GENIE registry, as well as members of the consortium for their commitment to data sharing. Interpretations are the responsibility of study authors. The authors also acknowledge our biopharmaceutical collaborators for sharing in our vision, their material support, and for their collaboration. Study data were derived using the PRISSMM™ system licensed and enhanced by Memorial Sloan-Kettering Cancer Center, Memorial Hospital for Cancer and Allied Diseases, and Sloan-Kettering Institute for Cancer Research (collectively ‘MSK’). Original system and improvements © 2019-2022 Dana-Farber Cancer Institute, Inc. Additional functionality and enhancements © 2023 MSK. All rights reserved. Memorial Sloan-Kettering Cancer Center, MSK, PRISSMM, and all associated logos are trademarks™ or registered® trademarks of MSK. The PRISSMM™ system or any portion thereof may not be reproduced, distributed, altered, or used in any manner without the prior written consent of MSK, the exclusive licensor of the PRISSMM system. Study data were collected and managed using REDCap electronic data capture tools hosted at each participating institution.9,10 REDCap (Research Electronic Data Capture) is a secure, web-based software platform designed to support data capture for research studies, providing (i) an intuitive interface for validated data capture; (ii) audit trails for tracking data manipulation and export procedures; (iii) automated export procedures for seamless data downloads to common statistical packages; and (iv) procedures for data integration and interoperability with external sources.9,10
References
- 1.Micheel C.M., Sweeney S.M., LeNoue-Newton M.L., et al. American Association for Cancer Research Project Genomics Evidence Neoplasia Information Exchange: from inception to first data release and beyond-lessons learned and member institutions’ perspectives. JCO Clin Cancer Inform. 2018;2:1–14. doi: 10.1200/CCI.17.00083. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Pugh T.J., Bell J.L., Bruce J.P., et al. AACR Project GENIE: 100,000 cases and beyond. Cancer Discov. 2022;12(9):2044–2057. doi: 10.1158/2159-8290.CD-21-1547. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.AACR Project GENIE Consortium AACR Project GENIE: powering precision medicine through an international consortium. Cancer Discov. 2017;7(8):818–831. doi: 10.1158/2159-8290.CD-17-0151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Consortium TAPG Release 16.1-public. Published online 2024. Available at. Accessed December 27, 2024. [DOI]
- 5.Biopharma Collaborative Core Team Annual Meeting 2022: AACR Project GENIE Delivers More Treasure. Cancer Research Catalyst. https://www.aacr.org/blog/2022/07/27/annual-meeting-2022-aacr-project-genie-delivers-more-treasure/ Available at. Accessed December 27, 2024.
- 6.AACR Project GENIE Begins Five-year Collaborative Research Project with $36 Million in New Funding American Association for Cancer Research (AACR) https://www.aacr.org/about-the-aacr/newsroom/news-releases/aacr-project-genie-begins-five-year-collaborative-research-project-with-36-million-in-new-funding/ Available at.
- 7.Institute of Medicine (US) Extending the Spectrum of Precompetitive Collaboration in Oncology Research: Workshop Summary. National Academies Press (US) 2010. http://www.ncbi.nlm.nih.gov/books/NBK210036/ Available at. [PubMed]
- 8.Institute of Medicine (US) Roundtable on Translating Genomic-Based Research for Health Establishing Precompetitive Collaborations to Stimulate Genomics-Driven Product Development: Workshop Summary. National Academies Press (US) 2011. http://www.ncbi.nlm.nih.gov/books/NBK54322/ Available at. [PubMed]
- 9.Harris P.A., Taylor R., Thielke R., Payne J., Gonzalez N., Conde J.G. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377–381. doi: 10.1016/j.jbi.2008.08.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Harris P.A., Taylor R., Minor B.L., et al. The REDCap consortium: building an international community of software platform partners. J Biomed Inform. 2019;95:103208. doi: 10.1016/j.jbi.2019.103208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Biostatistics: Project GENIE BPC: Genomics Evidence Neoplasia Information Exchange BioPharma Collaborative | Memorial Sloan Kettering Cancer Center. https://www.mskcc.org/departments/epidemiology-biostatistics/biostatistics/project-genie-bpc-genomics-evidence-neoplasia-information-exchange-biopharma-collaborative Available at.
- 12.Lavery J.A., Lepisto E.M., Brown S., et al. A scalable quality assurance process for curating oncology electronic health records: the project GENIE biopharma collaborative approach. JCO Clin Cancer Inform. 2022;6 doi: 10.1200/CCI.21.00105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Cerami E., Gao J., Dogrusoz U., et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2012;2(5):401–404. doi: 10.1158/2159-8290.CD-12-0095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Lavery J.A., Brown S., Curry M.A., Martin A., Sjoberg D.D., Whiting K. A data processing pipeline for the AACR project GENIE biopharma collaborative data with the {genieBPC} R package. Bioinformatics. 2023;39(1) doi: 10.1093/bioinformatics/btac796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Derry J.M.J., Mangravite L.M., Suver C., et al. Developing predictive molecular maps of human disease through community-based modeling. Nat Genet. 2012;44(2):127–130. doi: 10.1038/ng.1089. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.de Bruijn I., Kundra R., Mastrogiacomo B., et al. Analysis and visualization of longitudinal genomic and clinical data from the AACR Project GENIE biopharma collaborative in cBioPortal. Cancer Res. 2023;83(23):3861–3867. doi: 10.1158/0008-5472.CAN-23-0816. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Choudhury N.J., Lavery J.A., Brown S., et al. The GENIE BPC NSCLC Cohort: a real-world repository integrating standardized clinical and genomic data for 1,846 patients with non-small cell lung cancer. Clin Cancer Res. 2023;29(17):3418–3428. doi: 10.1158/1078-0432.CCR-23-0580. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kehl K.L., Lavery J.A., Brown S., et al. Biomarker inference and the timing of next-generation sequencing in a multi-institutional, cross-cancer clinicogenomic data set. JCO Precis Oncol. 2024;8 doi: 10.1200/PO.23.00489. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kehl K.L., Elmarakeby H., Nishino M., et al. Assessment of deep natural language processing in ascertaining oncologic outcomes from radiology reports. JAMA Oncol. 2019;5(10):1421–1429. doi: 10.1001/jamaoncol.2019.1800. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kehl K.L., Xu W., Gusev A., et al. Artificial intelligence-aided clinical annotation of a large multi-cancer genomic dataset. Nat Commun. 2021;12(1):7304. doi: 10.1038/s41467-021-27358-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Site-Specific Data Item (SSDI) Manual v3.1. Available at https://www.naacr.org/wp-content/uploads/2023/10/Site-Specific-Data-Item-SSDI-Manual_printed.pdf. Accessed December 27, 2024.