An Ontology-Based Information Management System for Pharmaceutical Product Development

Please download to get full document.

View again

of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report
Category:

Medicine, Science & Technology

Published:

Views: 4 | Pages: 10

Extension: PDF | Download: 0

Share
Related documents
Description
An Ontology-Based Information Management System for Pharmaceutical Product Development
Transcript
  An Ontology-Based Information Management System for Pharmaceutical Product Development Leaelaf Hailemariam 1 , Chunhua Zhao 2  , Girish Joglekar  1 , David Whittinghill  3 , Ankur Jain 1 , Venkat Venkatasubramanian 1 , Gintaras V. Reklaitis 1 , Kenneth R. Morris 4 , and Prabir K. Basu 5  . (1) School of Chemical Engineering, Purdue University, Forney Hall of Chemical Engineering, 480 Stadium Mall Drive, West Lafayette, IN 47907-2100, (2) Bayer Technology and Engineering (Shanghai) Co. Ltd, Shanghai, 201507, China, (3) Information Technology at Purdue, Ernest C Young Hall, 302 Wood Street, West Lafayette, 47907, (4) Industrial and Physical Pharmacy, Purdue University, 575 Stadium Mall Drive, West Lafayette, IN 47907-2088, (5) Purdue University, 315 Hovde Hall, 610 Purdue Mall, West Lafayette, IN 47907-2040     Abstract The development of pharmaceutical products and processes involves laboratory scale, pilot plant scale and commercial scale manufacturing. Through these steps, data on synthesis routes, material properties, processing steps, and scale up parameters are collected and used. Recent advances in process analytical technologies resulted in a large volume of heterogeneous data, which requires a systematic model of the associated information for optimal use. In addition, software tools are used to support decision-making and process modeling during process development. However, each tool creates and uses information in a specific form, making connection difficult amongst the tools and allowing little interaction with experimental data. In this work, an integrated information management system based on structured information is presented. This system is built to explicitly specify important concepts like physical properties and experiments using ontologies. Based on the developed ontologies, an information access and repository system is developed to allow convenient access to the repository for both the user and software tool. The infrastructure, designed to be accessible from the web, would allow simple search and browse as well as complicated queries, which can only be efficiently answered through the use of semantics in the information. This structure is extended to consider integration with software tools through the definition of an application-agnostic medium. Such an integrated information infrastructure provides a systematic approach to managing, sharing and reusing information, and is expected to result in considerable reduction in pharmaceutical process development time and better quality assurance. 1. Introduction Once a chemical entity, viable for its intended pharmaceutical application, is discovered, further development takes place in several integrated scales. These are the laboratory scale, pilot plant scale and commercial scales. At the laboratory scale, experiments are performed to characterize various synthesis routes and obtain process parameter values. Information generated at the laboratory scale can be used to improve manufacturing. At the pilot plant scale, studies are done to provide a detailed understanding of processing steps in the selected route and data needed for scale-up to commercial manufacturing is collected. At the  commercial scale, information related to manufacturing is applied in debottlenecking and productivity improvements. Problems identified at the manufacturing scale are communicated to the laboratory scale to identify root causes, and provide ideas on how to avoid similar problems in the future. In transitioning between these scales, several key activities have to be performed, such as, recipe synthesis, design and operation of the pilot plant, and scale-up to commercial production, as shown in Figure 1. The boxes indicate the major source of information for the given activities, with some overlap. Figure 1: Tasks in pharmaceutical product development and information flow between them Recipe synthesis and pilot plant design involve interaction with experimental data, for example, determination of material properties, while pilot plant design, operation and scale-up require use of software tools (e.g. ASPEN PLUS 2004 ®) which can also support decision making and process modeling in product development. In addition, there is information flow between these tasks; the software tools would make use of some experimental data (e.g. ASPEN PROPERTIES ®)and experimental analysis is assisted by software tools (e.g. computation of statistical trends). In addition, recent advances in process analytical technologies helped to provide a large volume of heterogeneous data, which requires systematic management for optimal use. The integrated use and management of heterogeneous information present several challenges. Each software tool   has a specific form of information creation and use, leading to islands of automation, which are difficult to connect and share functions among (Zhao et al., 2003). In addition, due to the lack of a versatile information model that can be used for data transfer, experimental data is likewise put in individual ‘silos’ as shown in Figure 2.    Figure 2: Use of information in a fragmented product development system  A large volume of heterogeneous data would require an efficient management system that allows for convenient access and retrieval as well as secure storage of information. The links between the pieces of available data (e.g. data from multiple experiments on the same sample) should be established in a systematic manner. Otherwise, pair-wise information translations would be required to cross these information divides, causing inefficiencies, uncertainties, costs, delays, and product quality concerns all along the product development pipeline. Thus, formal and explicit models of related information with well understood semantics (Schneider and Marquardt, 2002) are required. These information models should be easily accessible by humans and tools, and should provide a common understanding for information sharing There had been several attempts to address these challenges, including the Multi Dimensional Object Oriented Model MDOOM  (Burkett and Yang, 1995) and Process and Repository based  Approach for Requirements Traceability/ Chemical Engineering PROART/CE  (Jarke and Marquardt, 1996) for chemical process information integration associated with software tools. For laboratory information integration , one may cite  LIMS  (Paszko and Pugsley, 2000) and e-Lab Notebook (Zall, 2001). However, these solutions are frequently limited in their description of the relations between the different information entities and in their application to the pharmaceutical domain. The range of physical properties modeled was limited to fluid properties, while solid properties are of importance for pharmaceutical manufacture. In addition, integration of process information with experimental information was lacking. The objective of this research is to develop an integrated information management system that makes possible better data management, user view and edit capability of the data as well as efficient use of information in product development tools. This involves the development of an information structure, information storage systems, and interfaces for information exchange. The remainder of the paper is organized as follows. In Section 2, the development of the information model is described. Section 3 includes the use of the developed information model for laboratory information management. Section 4 describes how the information model may  be used to integrate experiments and tools. Conclusions and future directions are given in Section 5. 2. Development of the information model Several approaches may be used to model product development information. However, the information models are usually in closed form and only provide a limited view of the information (Zhao et al., 2005). A promising approach is the representation of information models by a structured, formal and well understood semantics (Schneider and Marquardt, 2002), otherwise called structured information . Without semantics, it would be difficult to organize related information and systematically manage meta-data (information about information) and allow integration of product development tools (Bayer et al., 2000). Ontologies, defined as “shared conceptualizations of a domain” (Gruber, 1993), are commonly used for modeling information in a formal and explicit manner. Compared to a database schema, which targets physical data independence, and an XML (eXtended Markup Language) schema, which targets document structure, an ontology targets agreed-upon and explicit semantics of information, and directly describes the graph structure consisting of the concepts  and their relations. In the pharmaceutical product development domain, several key concepts  need to be defined.  A central concept is the material  , which represents substances and mixtures (which are characterized by pure substances and their compositions). A material   has several physical and chemical  properties  (e.g. specific heat capacity), and can be involved in several experiments  (e.g. Compression Tests). Conversely, each material property value is obtained from a source of value, which includes experiments , literature  or mathematical models . The relationships between these concepts are shown in Figure 3. Figure 3: The relations between major concepts in pharmaceutical product development  The concepts defined above are described by a set of attributes. A material   is described by its composition, phase, properties, roles (e.g. does it help flow of a powder of which it is a part?) and experiments it participated in. A  property   is represented by the material   it belongs to, its value , the source  of that value, the conditions under which that value was measured (called its context  ). The description of an experiment   would include the materials involved, the property measured, the experimenter, location of the experiment, the date and time of experiment, the equipment used, the procedure followed and the experimental data obtained. (a) (b) Figure 4: (a) Protégé input Interface for creation of material properties (b) Graphical view of the relations between concepts Based on explicit expressions of the relations between the concepts described above, material, property and  experiment   ontologies were developed. Ontologies were also developed for ‘secondary’ level concepts like value, source, context and other descriptors as necessary. Web Ontology Language (OWL, 2004), was used in this work to encode ontologies. The ontologies were developed in close collaboration with colleagues in the Department of Industrial and Physical Pharmacy at Purdue University with several revisions during development. The visualization tools provided by the ontology editor, Protégé (see URL) and the plug-ins in the editor including the view of class hierarchy, the graph view (shown in Figure 4) were found to be very convenient in the collaboration. 3. Development of an integrated information infrastructure  An effective information management system should manage the large amount of experimental information and knowledge generated during experimental analysis and facilitate efficient decision-making. Users of the system should be able to conveniently upload, browse and search the available data. This requires the information management system to have the following basic components: an information model, an information repository and interfaces that can communicate with the user and allow the user to perform such tasks as search and browse. To that end, the developed ontology has been used as the foundation for the information repository (Zhao et al., 2005). Folder structures to store experimental files are created based on the concept hierarchy defined in the ontology. The system allows transition between related instances and search by keyword as well by hierarchy. A web-based interface
Recommended
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x