|
 |
CGL7 - Clinical Genomics Level Seven
Healthcare & Life Sciences
|
CGL7 is a research prototype to enable the exchange of clinical and personalized genomic data between interested parties. In many cases the exchange of genomic data is done between disparate organizations and standards are crucial to outline the use and exchange of the data in healthcare practice. This project follows the design objectives of the HL7 Clinical Genomics Standards Specification. HL7 focuses on the development of standards for the specification of messaging and documentation in healthcare. The main design principle of the HL7 Clinical Genomics Specifications is to enable the 'encapsulate & bubble-up' paradigm.
CGL7 is a set of web services that:
- Encapsulate raw genomic data
- Bubble-up the most clinically-significant items
- Link them with known clinical phenotypes found in publicly-available sources (Ontologies, papers, reference databases, etc.) and with observed clinical phenotypes in the patient, found in clinical repositories (e.g., an enterprise EHR system).
Our collaboration with Harvard-Partners Center for Genetics and Genomics (HPCGG) is an example of the use of CGL7. Patient samples are sent from one of Partners' hospitals with a request for specific genetic tests. The results are annotated by HPCGG geneticists and a report is sent back to the referring clinician. CGL7 is used to enrich the geneticists' annotations by accessing publicly-available ontologies like OMIM and search for the most up-to-date studies related to the testing results (e.g., the patient's somatic mutations found in a tumor tissue). The communication between the HPCGG lab, the enterprise clinical information system and CGL7 is done based on the HL7 Clinical Genomics specifications.
CGL7 main features
- Encapsulate raw genomic data into an HL7 Clinical Genomics message, including transformation services of proprietary data formats to standardized formats like MAGE and BSML
- Access patient's clinical history stored in an enterprise EHR system
- Access all major ontologies that provide genotype-phenotype relationships like OMIM, PharmGKB, etc.
- Parse the encapsulated raw genomic data and bubble-up selected genomic data items based on ontological knowledge as well as the patient clinical data
- Compare two Genotypes (the data model at the heart of the HL7 specs) in order to provide case-based reasoning services to decision support application that will use CGL7 as a specialized clinical genomics middleware
- Find a similar pedigree in case base of pedigrees in order to support risk assessment applications that base their assessment on family history
CGL7 Utilization
CGL7 web services can be used either in conjunction with IBM Clinical Genomics solutions (e.g., when a persistence repository is needed) or as a standalone service hosted by IBM or within an enterprise.
| |
|
|
