Time-Related Patient Data Retrieval for the Case Studies from the Pharmacogenomics Research Network
- PDF / 327,299 Bytes
- 6 Pages / 595.276 x 790.866 pts Page_size
- 116 Downloads / 177 Views
ORIGINAL PAPER
Time-Related Patient Data Retrieval for the Case Studies from the Pharmacogenomics Research Network Qian Zhu & Cui Tao & Ying Ding & Christopher G. Chute
Received: 18 June 2012 / Accepted: 9 October 2012 / Published online: 18 October 2012 # Springer Science+Business Media New York 2012
Abstract There are lots of question-based data elements from the pharmacogenomics research network (PGRN) studies. Many data elements contain temporal information. To semantically represent these elements so that they can be machine processiable is a challenging problem for the following reasons: (1) the designers of these studies usually do not have the knowledge of any computer modeling and query languages, so that the original data elements usually are represented in spreadsheets in human languages; and (2) the time aspects in these data elements can be too complex to be represented faithfully in a machine-understandable way. In this paper, we introduce our efforts on representing these data elements using semantic web technologies. We have developed an ontology, CNTRO, for representing clinical events and their temporal relations in the web ontology language (OWL). Here we use CNTRO to represent the time aspects in the data elements. We have evaluated 720 timerelated data elements from PGRN studies. We adapted and extended the knowledge representation requirements for EliXR-TIME to categorize our data elements. A CNTRObased SPARQL query builder has been developed to customize users’ own SPARQL queries for each knowledge representation requirement. The SPARQL query builder has been evaluated with a simulated EHR triple store to ensure its functionalities. Keywords Temporal ontology . Semantic web . Pharmacogenomics studies . SPARQL query builder Qian Zhu and Cui Tao contributed equally. Q. Zhu : C. Tao (*) : C. G. Chute Mayo Clinic, Rochester, MN, USA e-mail: [email protected] Y. Ding Indiana University, Bloomington, IN, USA
Introduction Pharmacogenomics deals with the influence of genetic variation on drug response in patients by correlating gene expression or single-nucleotide polymorphisms with a drug’s efficacy or toxicity [1]. The Pharmacogenomics Research Network (PGRN [2]) is a collaborative partnership of research groups funded by the U.S. National Institutes of Health to investigate pharmacogenomics studies. Huge volume of pharmacogenomics data including a large portion of time related data elements is accumulated across this network. We presented the study of harmonization and semantic annotation of the pharmacogenomics data from PGRN network previously [3], but we have not yet focused the associated temporal issues. Temporal information is very important in pharmacogenomics studies. For example, investigators need to track patients’ medical history in order to find the drug effects for genetic variations during a specific time frame. Therefore, there is an urgent need to capture the temporal information not only for the purpose of representing these data elements in a standard way for interoperability a
Data Loading...
