Case Study |

Connected Patients

Context

The client has developed a patient facing platform that created a space for patients to discuss their specific disease and experience with others going through similar experiences. The goal was to utilize this platform to form a collective picture of patient health and gather RWD (Real World Data). Their platform collected patient reported data, medical records, medications, clinical care, comorbid diseases, and laboratory and NGS data.

The Project

We designed their knowledge management platform specifically to use for ontologies, mapping, and NLP projects.  It needed to incorporate automatic workflows, release cycles, and user management. Mapping was geared towards immunologic and blood-based metabolites. In order to perform NLP, a metadata and annotation structure was also implemented. Through NLP, we were able to take this phonotypic data and overlay is with existing genetic information to evaluate any trends along with compare with open information sources. The data was collected to be implemented under FAIR (findable, accessible, interoperable, and reusable) data principles. This project was completed by a two person Arrayo team alongside our client’s data team.

We advised and provided further Knowledge Engineering services:

  • Specify Minimal Attributes Set for Entities (MASE): ID syntax and functionality
  • Specify strategy and architecture for fundamentals such as: ID Management System, Ontology Management System, and Knowledge Management System
  • Prepare ontologies for release into public domain
  • Quality Control, generate and edit ontologies
    • Considering content, domain coverage, and appropriateness s OMS backbone
      • Unified Medical Language System (UMLS)
      • BioPOrtal
      • OBO Foundry
    • Text mining/ NLP
      • Linguamatics
      • SCiBite

We addressed change management with Presentations and Courses:

  • Ontologies 101: building principles
  • Protégé presentation
  • Client specific ontology training: strategy, conditions, building blocks, etc
  • Minimal Attributes Set for Entities (MASE) training
  • Unified Medical Language System (UMLS) training
  • Spotfire template training to understand visualization interface

We redesigned and standardized:

  • Bioloader functionality and workflow redesign
  • Chemical compound descriptors standardization and normalization

Outcome

  • Our client company helped connect people who have similar health conditions.
    • Arrayo helped our client company design a business and integrated data models to boost precision medicine
    • Helped our client monetize their data and increase corporate valuation
    • Created Spotfire templates and views to support business development and partnership efforts
    • Designed an educational framework to promote adoption of the overall data strategy between our client and a complex partnership
    • Successfully delivered change management function across a complex 4 company merger.