-
Notifications
You must be signed in to change notification settings - Fork 45
Home
ONCOSCAPE is an open-source, web-accessible visualization tool created at the Fred Hutchinson Cancer Research Center (Fred Hutch) under the auspices of the Seattle Tumor Translational Research initiative to be utilized by researchers and clinicians.
Oncoscape is a data visualization platform that empowers researchers to discover novel patterns and relationships between clinical and molecular data. Through a suite of interoperable tools, Oncoscape offers a unique and intuitive approach to hypothesis refinement. Both CRAN and Bioconductor offer numerous R software packages for handling large scale biomedical data using open source statistical analysis and visualization methods. Similarly, AngularJS in the browser provides a rich and nimble user experience.
Unique features and contributions of Oncoscape are based on the intersection of cancer biology, clinical oncology, and advanced computational methodologies. Through the utilization of sophisticated analytics, domain specific knowledge, and curation of clinical records, novel visualizations and computational algorithms were developed to connect and display informative combinations of diverse data types. The ability for the user to then selectively link results among methods through simple browser interactions allows for rapid hypothesis exploration without the need for computational skills. Additional and external methods and features can be easily integrated into Oncoscape through the loosely coupled communication protocol, which allows for unimpeded growth and adoption of emerging technologies.
Deidentified data comes from public resources such as The Cancer Genome Atlas (TCGA). The TCGA is a coordinated effort of the NCI and the National Human Genome Research Institute (NHGRI) to accelerate our understanding of cancer biology through large scale, high-throughput genomic analysis. Combining multiple datasets allows users to both explore patterns and connections among molecular and clinical characteristics then validate findings in an independent study. Finding consistent trends across different datasets can help researchers define new testable hypotheses leading to new discoveries relevant to patient care.
However, small, disjointed datasets may have too few patients to properly compare cancer subtypes, and not being able to distinctly define the patient group may dilute the analysis. It is the intention of Fred Hutch to allow contributors to pool large datasets on cancer patients. This will give researchers the statistical power to explore populations of a sufficient size so they can detect patterns among patient subgroups, e.g. triple negative breast cancers.
This public forum also allows the external coding community to contribute their own expertise, leading to broader support and maintenance for more flexible, extensible, and transparent software development. The open source Oncoscape repository aims to widen the playing field so computational experts with various skill sets can enhance the tools and software currently used by cancer biologists and oncologists.
Oncoscape