Skip to content
Tiffany J. Callahan edited this page Oct 28, 2019 · 21 revisions

PheKnowVec


Collaborators:
Informatics Team
Sara Deakyne Davies, Michael G. Kahn, Lawrence E. Hunter

Clinical Team

Translational Research Team
Adrianne L. Stefanski, Nicole Vasilevsky, Xingmin Aaron Zhang, Peter N. Robinson


Project Description:
Computational phenotyping (CP) approaches have great potential to aid in diagnosis, prognosis, therapeutic decision-making, and identification of mechanisms or novel biomarkers. Current CP methods are unable to solve three CP barriers:

  1. CP definitions may have limited generalizability because they are tailored to specific source vocabularies or hospital systems.
  2. CP definitions may lack translational relevance because they primarily rely on clinical data requiring additional mapping to incorporate, for example, molecular or physiologic data.
  3. CP definitions may lack scalability because the current process for creating definitions is a time-consuming, iterative process requiring both domain expertise and external validation.

PheKnowVec is a novel method for deriving, implementing, and validating CPs that addresses these barriers by:

  • Mapping standardized clinical terminology concepts to linked open data.
  • Using embedding methods, which convert large complex heterogeneous data into scalable compressed vectors without semantic information loss.

PheKnowVec leverages standardized clinical terminologies and open biomedical ontologies to derive, implement, and validate CP definitions in a scalable embedded structure. The scalability of PheKnowVec CP definition embeddings improve the process of implementing and validating existing and/or novel CPs. The use of biomedical ontologies significantly facilitates the integration of relevant external data.



Publications and Presentations