Week 9 March 29, 2011: Translational Medicine Ontology and Knowledge Base - from the perspective of semantic technologies

Printer-friendly version

This week we will discuss the Translational Medicine Ontology and Knowledge Base paper.

Revised TMO Paper [Download]

titanpad for today

Topics to discuss:

(a) The problems addressed in the paper


(b) The Semantic technologies that were used


(c) Discussion of the benefit of semantic technologies in grounded example showing something that would be hard to do without semantic technologies.
(e) claims that you expect to include in your write up.


(f) roles and responsibilities for each team member

Reading Assignment for this class:

  • Identify some semantic technologies you'd like to discuss (list the ones in the tmo to start with)

Written Assignment for this class:

  • Update this website with topics and references where applicable
Name Topic for Discussion Section and Page References
Eric
  • What will be the benefits to incorporating data provenance into the ontology and knowledge base? The authors claim that including provenance will enable interoperability with large scale eScience work. I'd like to discuss evidence for this statement.
  • How might the authors test the scalability of this system, considering they used only 7 synthetic patient records? EHR systems such as the SMart platform offer there own services to access patient health data in an RDF format. Is this system still usable in a context where patient health record data is gathered from an external source, such as the SMart API?
  • How does the system handle errors in the data being processed? Is this being accounted for in any way? What could the ontology and knowledge base incorporate to ensure data quality (hint: provenance, instance validation, etc.)?
  • Conclusion, pg. 20
  • Data Sources, pg. 11
Jin
  • Automatic inferencing and consistency checking is one of the key feature provide by OWL. What are the types of inferences and consistency checking are supported by TMO? Why such inferences and consistency checking are important and interesting?
  • The paper demonstrates the usefulness of Semantic Web technologies in the integration of heterogeneous data related to health, and provide a knowledge base that is helpful in answering various health related questions. The questions are answered using SPARQL queries through a sparql endpoint. However, considering the patients and health specialists may not have much knowledge about SPARQL and Semantic Web technologies, therefore, a user-friendly front-end interface is crucial in the senses. I would like to discuss what are the choices of front-end design that can simplifies the users experience of using the knowledge base? How? What are the limits of such design?
  • Also, the scalability problem is also important. Considering if we have web-scale health data, how TMKB can handle such amount of data w.r.t both queries and inference?
  • Ontology Design(page 10), TMO,
  • SPARQL Queries(page 15,16,17,18)
  • Data Source(page 11)
Ping
  • I didn't find individuals in the TMO ontology. How do you combine the ontology and data?
  • What data format/standard do you use to create the the synthetic patient records?
  • In Figure 3. TKMB overview, at the bottom, there are six data sources. How do you utilize these data sources in TMKB?
  • Ontology Design, pg.10
  • Data Sources, pg. 11
  • Figures, pg. 28

Presentation Assignment for this class:

  • None, we will discuss in class