Clinical and Molecular Informatics Analysis:
It is our intent to create a new support service within biomedical informatics that can support data mining for relationships between clinical disease phenotypes and underlying genomic heterogeneity, as defined through genetic phenotyping, genomic expression profiling, proteomics, or cellular studies. A project to create an ontology and structure for description of clinical diseases is below (Phenotype Analysis). Essentially, establishment of clear clinical phenotypes and associated disease variation becomes the starting point for the very complex problem of relating specific disease states to underlying genetic and genomic variations. Most diseases today are recognized as complex traits where interactions between a set of inherited genes and the environment produce recognizable physiologic variation, including extreme variations that are recognized as named diseases. However, even with diseases caused by single gene mutations, e.g., sickle cell disease, disease expression (at the level of associated medical problems) could be altered by other disease-modifying genes. Hence, one needs the ability to store large data sets that describe molecular phenotypes of individuals and then relate those data to clinical diseases. Moreover, in clinical research, data are obtained under various experimental conditions. We will use an ontology for describing not only the variables, but also conditions under which data were collected.
Bleeding History Phenotype Analysis, Questionnaire, and Electronic Data Recording Instrument:
As part of the Rockefeller University Center for Clinical and Translational Science phenotyping initiative, the Bleeding History Ontology (BHO) was developed as an application ontology devoted to the standardized recording of phenotypic data related to hemorrhagic disorders. It was constructed with the goal of homogenizing domain terminology related to bleeding history phenotypes, as well as regulating questionnaire instruments for studies related to various inherited, acquired, and drug-induced hemostatic disorders. The ontology will be used to:
- develop clinical and research instruments for collection of phenotype data in a consistent manner
- support data analysis collected by such tools
- perform meta-analysis by annotating and aggregating data with and/or from non-standardized questionnaires
The BHO was constructed from the detailed bleeding history questionnaire developed by Dr. Barry Coller, which was based on his research and clinical experience.1 The ontology was created with the assistance of Protégé, and is currently implemented in the Protégé Frames format. The ontology consists of two parts: the domain and the application. The domain section is best described as terminology and concepts pertinent to event history of bleeding and bleeding disorders. The application portion contains detailed information relating the questions and answers in the questionnaire and is fully cross-referenced to the classes within the domain section. Entities within the ontology are annotated with cross references to the Unified Medical Language System (UMLS), and where appropriate, to Online Mendelian Inheritance in Man and ICD-10.
The two main functions of the BHO will be to assist researchers in constructing their own questionnaires regarding hemorrhagic disorders and to provide support for data analysis. The Ontology will provide guidance and uniformity in data aggregation, questionnaire inter-operability, and consistency in bleeding phenotype records from across multiple disciplines. More specifically, the ontology will contain explicit examples of questions that have been used in previous questionnaires and how answers to those questions map into the domain.
In order to assist in data analysis for phenotype studies that focus on bleeding disorders, the BHO can support data abstraction, interpretation, and integration, with the aid of Protégé or other, custom built software.
Currently the BHO, the Bleeding History Questionnaire (BHQ), and the electronic data recording instrument are complete and will soon be deployed for clinical studies and made available to the entire scientific community through the National Center for Biomedical Ontologies (NCBO). If you have comments or suggestions, or if you are interested in utilizing this Ontology in your work, please contact: Mr. Nickolay Khazanov, Dr. Barry Coller, Dr. Andreas Mauer, or, Mr. Edward Barbour.
- Coller BS, Schneiderman P. Clinical evaluation of hemorrhagic disorders: The bleeding history and differential diagnosis of purpura. In: Hoffman R, Benz EJ, Shattil SJ, Furie B, Cohen JJ, eds. Hematology: Basic Principles and Practice. New York: Churchill Livingstone; 2007:In Press.