What is the Xena Project?
The University of California Santa Cruz (UCSC) created the UCSC Xena Browser to securely analyze and visualize private functional genomics data sets in the context of public and shared genomic/phenotypic data sets. Xena enables the use of big genomics data resources (i.e., TCGA) by the broader scientific community, supporting dynamic queries, exploration, and visualization of these data.
Why collaborate with TIES?
Xena had a challenge: researchers wanted to select samples based on phenotypic features that were detailed only in pathology reports. Through the NCI ITCR (U24CA180921-04) grant, Xena worked with TIES to develop its software to expose the TCGA pathology reports stored in the TIES repository using a REST API. The TIES natural language processing-based intelligent clinical data search, within the UCSC Xena Browser, made it possible for genomics researchers to use clinical report text (such as the pathology reports) to find specific samples and genomic results associated with particular phenotypes. In other words, TIES helped automate the natural language processing and semantic searches, with controlled vocabularies and ontologies, to parse through pathology reports associated with the genomic data set the users define using the Xena Browser. These narrowed searches allowed researchers to explore genomic features on only the samples that met their detailed phenotypic requirements.
How did the system develop?
This collaboration was the product of several months of iterative design sessions between the two teams.User Experience Designers interviewed real users from the Xena community to determine where natural language processing of clinical report text could most help in the selecting of samples to view. From there, the TIES team developed the first versions of the REST API services, based on concept or free text searching and document retrieve of the TIES reports. These services were then tested and incorporated within the Xena brower.
What comes next?
The hope of this collaboration is for Xena to provide an example that can be emulated by other portals, potentially leveraging the same open source software that was developed. The completed project also warranted a lot of knowledge for the TIES team, which they will use when redeveloping the new TIES interface and software to better serve researchers and the scientific community.
For a demonstration, go to the following link: TIES-Xena Demo. The data displayed throughout the TIES-Xena demo is all publically available TCGA data.