1 Current status of ontologies in Biomedical and Clinical Informatics Rishi Kanth Saripalle University of Connecticut, Storrs [email protected] Abstract It is becoming an impossible task for managing research data in the field of biomedical and clinical informatics. These huge data sets cannot manually be analyzed, interpreted or processed to acquire inferred knowledge efficiently. We need intelligent agents or computer systems to help us in doing these tasks and hence it becomes mandatory to represent medical knowledge in computer process able format. Semantic technology and ontology can be used to partial solve the data management problem in medical informatics. Semantic knowledge representation allows the intelligent agents or computers to interpret the data and acquire inferred knowledge. Hence, ontology design is an important aspect of medical informatics, and reusability is a key issue that is determined by the level of compatibility among ontology concepts and among the theories of the biomedical domain. In this paper we will discuss the role of ontologies in representing medical informatics. First, we will talk about the principles of ontologies and few examples. Next we will study some fully developed biomedical ontologies and their differences. Finally, we study the applications of ontologies in biomedical research. Terms: Ontology, biomedical ontologies, medical informatics, clinical informatics, clinical decision support systems, knowledge representation, patient health record (PHR), clinical trials, Protégé, SNOMED-CT, GALEN, Gene Ontology (GO), Ontology integration, Ontology Mapping, NCodes. I. Introduction In the last few decades much research has been done in the field of biomedical informatics and voluminous amount of research data has been collected in the fields of clinical research, biomedical research, life sciences, gene research, patient records, clinical trials etc. Simultaneously various biomedical tools have been developed to perform wide range of function like data mining, data management, data collection etc. This in turn has forced scientists to analyze and structure the knowledge to make further inferences from the present knowledge. A survey conducted showed that DNA sequence databases have been doubling for every 18 months. Most of the existing databases have overlapped data as they are built independent to each other. Database systems today are facing the task of serving ever increasing amounts of data from growing complex user community which is getting more and more demanding. All the researchers need almost the same data but with different meaning or context. This makes semantics very important for this domain. II. Ontology Ontology - “science of being" - typically has different meanings in different contexts. The word originated in philosophy where several philosophers - from Aristotle (4th century) to Leibniz (1646-1716), and more recently the 19th Century major ontologists like Bolzano, Brentano, Husserl and Frege have provided criteria for distinguishing between different kind of objects and the relations among them. The objects can be both concrete and abstract. In the late 20th century, Artificial Intelligence (AI) adopted the term and began using it in the sense of a "specification of a conceptualization" in the context of knowledge and data sharing [1]. 2.1 Definition Ontology is “hierarchal structuring of knowledge about concepts by sub-classing them2according to their properties and qualities” [1]. It can also be defined as “a declarative model of a domain that defines and represents the concepts existing in that domain, their attributes and the relationships between them” [1] [2]. Ontology gives the description of concepts and the relations that can exist between them. The concept is very important for data sharing and knowledge representation. 2.2 Classification Ontology can be classified according to level of detailed knowledge they provide. Upper Ontologies provides very generic knowledge with low domain specific knowledge. For example, Disease ontology is upper ontology compatible for any biomedical domain. General ontologies represent knowledge at an intermediate level of detail independently of a specific task. Domain ontologies represent knowledge about a particular part of the world, such as medicine, and should reflect the underlying reality through a theory of the domain represented. For example, Gene Ontology, Finally, ontologies designed for specific tasks are called application ontologies. 2.2 Description We have defined ontology as specification of concepts and relation between them. In ontology, concepts of the domain are represented by “classes”. The features and attributes of the concept are described by “properties or slots”. Together with “instances” which are individual of a class it constitutes the knowledge base of the domain. Classes are the main focus in ontology. Classes can be sub-classed to describe more specific features of a class. For example, if we define a class Wines, it includes all the wine classes in the wine domain. The wine class can be sub-classed to specify more specific wines like Red Wine, White Wine etc. Instances are individuals related to a same class. For example, Australian Yellow Tail is an individual for Wine class. Slots or Properties can be created to describe properties of a class or instance. For example, we can define a property named “Has_Color” which holds the color of a particular wine class or instance. The figure 1 shows the summary of the above discussion. Therefore in particular describing a domain in ontology includes [3]: 1. Defining the concepts of the domain as classes. 2. Defining the individuals of the class as instances. 3. Defining the attributes of the individuals as properties 4. Filling the properties values for the instances. 2.3 Advantages: Ontologies are developed and defined to share the knowledge among the researchers working on the same domain. The main reasons for developing ontology are [3][4]: 1. Sharing the knowledge in the same domain is one of the common goal for which ontologies are developed. For example, many websites provide medical information about various concepts in the medical domain. If they use the same medical terms for describing the information, the data can be integrated easily from different sources by the computer agents and