Precision medicine requires an end-to-end learning health care system, wherein the treatment decisions for patients are powered by the prior experiences of every similar patient who has preceded them. Oncology is currently leading the way in precision medicine because the factors that fuel cancer initiation, progression, and recurrence—namely, the genomic and other molecular characteristics of patients and their tumors—are routinely collected at scale. A major challenge to this approach, however, is that no single institution is able to sequence and treat sufficient numbers of patients to improve clinical decision-making independently. The AACR Project GENIE® (Genomics Evidence Neoplasia Information Exchange) registry was established in 2015 to address this challenge and realize the promise of precision medicine.
AACR Project GENIE® is an open-source, international, pancancer registry of real-world data assembled through data sharing between 19 leading international cancer centers. The registry leverages ongoing clinical sequencing efforts at participating cancer centers by pooling their data to serve as an evidence base for the entire cancer community. The consortium and its activities are driven by openness, transparency, and inclusion to ensure that the project output remains accessible to the global cancer research community and ultimately benefits patients.
In its seventh year, AACR Project GENIE® took several bold steps toward delivering on the promise of precision medicine:
This fall, the GENIE dataset was put to a unique use in a global genomic data hackathon organized by the Children’s Tumor Foundation. The event—titled Hack4NF—provided researchers, data analysts, genomic experts, computational biologists, statisticians, and health care startups with access to de-identified patient data from Project GENIE and two other registries and challenged the participants to develop solutions to the challenges faced by neurofibromatosis researchers. AACR Project GENIE® data powered multiple projects during the hackathon, including one of the prize-winning ideas. The winning team treated the mutated genes in a tumor sample like a sentence and used Natural Language Processing to create a model to predict a tumor’s cancer type. The model could be used to identify neurofibromatosis tumor subtypes and to explore their relationship with the cancers that most typically arise from neurofibromatosis.
Public Data Releases
≤ 18 @ sequencing
Young Adult Patients*
≥ 18 ≤ 39 @ sequencing
Major Cancer Types*
Unique Cancer Subtypes*
*12.1 public release.