Discover the new sample query and subscription and the improved user interface workflow
The Hyve has just released a new version of Glowing Bear, the user interface that was specifically developed for the tranSMART 17.1 data warehouse. The workflow in Glowing Bear 2.0 has become more user-friendly. Moreover, an important improvement is cohort selection based on information related to laboratory samples such as tissue samples or sequencing data.
While Glowing Bear was officially launched last September, the Hyve and its collaborating partners never considered it a finished product. Now, we are ready to release Glowing Bear 2.0. In this blog we want to discuss changes we made to improve the workflow and added functionality to improve the query search.
Demo video of the new features in Glowing Bear 2.0
Sample query and subscription
In the first version of Glowing Bear, researchers would create a cohort by selecting the variables of interest: gender, age, diagnosis, treatment, et cetera. This means that they always start with the total number of patients, with the search narrowing down on the cohort of interest with every subsequent variable.
In Glowing Bear 2.0 it is still possible to select a cohort starting with the entire patient population, but The Hyve’s team also created the option to select a cohort based on information in tranSMART that is available for patients on Diagnosis, Biosources or Biomaterials. This search can be further narrowed down by selecting patient characteristics such as age and gender. This functionality was developed in collaboration with the Prinses Máxima Center for Pediatric Oncology in Utrecht and Health-RI consortium.
At the Prinses Máxima Center, the Diagnosis section will contain a range of tumor types and other tumor related data (e.g. topography). The biosource section contains information on biological samples collected from a patient: blood tests, urine tests and tissue samples. The biomaterial section could contain DNA and RNA sequences with information on the type of methodology used or which analysis was performed. The contents of these sections can be customized to the needs and requirements of any given hospital or research institution. For example, the image section can be added that contains information on X-ray, MRI or CT images.
To give you an idea how this sample based query works: the illustration below shows a query for the number of biomaterials with "genomic DNA" gave 3 subjects that matched the selection, as the summary at the top indicates. The total number of “genomic DNA” found was 4: for one subject two samples were available. To retrieve a list of these biomaterials, an export of the query needs to be made in the Data Export tab.
In addition, it is now possible to subscribe yourself to a saved cohort that contains sample information and not only patients as before. Subscription means that you will receive an email when there are changes either concerning the subjects or the samples that meet the saved criteria.
Current users of Glowing Bear will notice we made some changes in the 2.0 version to make the user interface more user-friendly and improve the workflow. For example, the tabs on top of the page labelled Data Selection, Analysis and Export, have been renamed Cohort Selection, Analysis and Data Export.
The tabs still link to the same pages, so it’s only a minor change. The first tab is still used to define patient groups based on inclusion and exclusion criteria. These criteria can be typed into a search bar, selected from a drop-down menu, or added by drag-and-drop.
Saved searches could previously be found under the header Queries in the left sidebar. The name of this header has been changed to Cohort as medical staff is more familiar with this term than with the IT-term ’query’.
In the Cohort Selection page, a panel has been added with Variables that lists the variables available for a specific cohort, for example age, heart rate and tumor type. The variables can be displayed in a tree view or a category view.
The third tab, Data Export, still offers the option to export data subsets from the tranSMART data warehouse for in-depth analysis with programming languages such as Python or R.
In addition, you can select the variables of interest by checking the boxes of the Variables views in the Left panel, and the selection is displayed in the data table. These features were present before in the same tab for performing cohort selection. Moving them to Data Export, provide to the user a better focus on the meaning of the tab.
Another feature we changed in Glowing Bear 2.0: it is now possible to select subjects groups both under the Cohort Selection tab and under the Data Export tab, using the saved cohorts panel on the left side. This creates a more streamlined workflow for experienced users, as they can now skip the first page and don’t need to switch tabs when selecting and exporting patient cohorts.
If you are already familiar with Glowing Bear, we hope you like the changes we made to the user interface and appreciate the visualisation options we added by incorporating the Fractalis platform. We welcome any feedback on Glowing Bear’s functionalities and encourage you to share ideas that could further improve the user interface.
So please, feel free to contact us or leave a comment.
- Glowing Bear officially launched
- Keycloak: for easy, secure and flexible user management
- TranSMART data exploration and analysis using Python Client and Jupyter Notebook
- Python and R client: data retrieval for analysis made easier