Model Overview
What is a data model?
A data model organizes data elements and standardizes how the data elements relate to one another. It explicitly determines the structure of the data.
The Gray Foundation’s data model is derived from several data standards such as the Genomics Data Commons but has also been adapted to fit the needs of the consortium. It outlines, defines, and standardizes how data such as clinical data are represented and how they relate to one another, e.g. a patient has a diagnosis and receives therapy. One of the most important relations is of clinical data to generated data -- in Gray Foundation, most generated data are human data and need to be tied to the original patients for useful analysis.
The section Clinical Data explains what clinical data are prioritized.
In the data model, attributes are grouped into “components” or “modules”, e.g. patient-related attributes such as age, sex, etc., are in a patient core component.
Attributes appear as columnnar fields in a table when collecting data. They may be required or optional and may have controlled terminologies for the values.
What is metadata?
Metadata is additional, standardized information included alongside the data to give it context—data about the data, if you will. Metadata is what allows data in the portal to be searchable, discoverable, accessible, re-usable, and understandable to others, including those who were not involved in the data generation process. Metadata can be descriptive (i.e., the name of the file), administrative (i.e., provenance information), or research-based (i.e., information about the sampling and handling of data).
-- AD Knowledge Portal Glossary
Metadata can also be thought of as "data about data", while clinical data can be thought of as "data about patients". On the Synapse platform, adding metadata to data entities (files) is most often called "annotating", and metadata is interchangeably called "annotations". The Dataset and File Metadata section goes into more detail what annotations are expected for datasets and different file types.
Support and contributions
For questions/discussions, suggestions, and issues (bugs) regarding the data model, it is preferred that members submit an issue at our source repository. Note that this requires a GitHub account. If you do not have a GitHub account, please reach out to one our DCC staff listed in Contacts.