An ontology for computer-aided modeling of cellular processes

Abstract

Cellular processes form the hardware layer of living organisms. Malfunctions in cellular processes are responsible for most of the currently incurable diseases. Not surprisingly, knowledge about cellular processes are growing at an enormous rate. However, today’s molecular biology suffers from lack of a formal representation system for cellular processes. Most of the knowledge is locked in literature, that are not accessible to computational analysis and modeling. Given the complexity of the system we are attacking, the need for a representation system and modeling tools for cellular processes are clear. In this dissertation, we describe an ontology for modeling processes. Our ontology possesses several unique features, including ability to represent abstractions and multiple levels of detail, cellular compartments and molecular states. Furthermore, it was designed to meet several user and system requirements, including ease of integration, querying, analysis and visualization. Based on this ontology we also implemented a set of software tools within the Patika project. Primary use cases of Patika are integration, querying and visualization, and we have obtained satisfactory results proving the feasibility of our ontology. Compared with existing alternative methods of representing and querying information about cellular processes, Patika provides several advantages, including a regular representation system, powerful querying options, an advanced visualization. Moreover Patika models can be analyzed by computational methods such as flux analysis or pathway activity inference. Although it has a more steep learning curve compared to existing ad hoc representation systems, we believe that tools like Patika will be essential for molecular biology research in the future.