Briefings in Bioinformatics Advance Access originally published online on May 11, 2006
Briefings in Bioinformatics 2006 7(2):207; doi:10.1093/bib/bbl011
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Book Reviews |
Ontologies for Bioinformatics (Computational Molecular Biology).
Kenneth Baclawski and Tianhua Niu The MIT Press; ISBN: 0-262-02591-4; Hardcover; 440pp; 2005; £29.95. As the amount and variety of available knowledge in the biomedical domain grows, we need mechanisms to make our results and conclusions available in an explicit and structured machine-accessible form. Ontologies are viewed as a means to achieve this goal. There is much excitement in the scientific community about ontologies and how best to utilize them. ‘Ontologies in Bioinformatics’ attempts to address this need for people who have no prior background and who need an introduction to the field.
This book is organized into three sections: providing a broad overview of what an ontology is, how to (programmatically) use it, and how to construct an ontology. There is also an introduction to the reasoning methods and their relationship to ontologies. The first section ‘Introduction to Ontologies’ describes hierarchies and their meanings, gradually leading into XML. It also touches on rule engines and theorem-proving software. The section ends with chapters on the Semantic Web and a brief survey of ontologies in bioinformatics.
The second section ‘Building and Using Ontologies’ focuses on information retrieval, query languages and transformations using XML and Perl. The section ends with a chapter on building bioinformatics ontologies, and it provides pointers to ontology building tools. The third section discusses reasoning methods such as inductive, deductive as well as Bayesian reasoning and their connections with ontologies. The section ends with a chapter on the Bayesian Web, describing a stochastic extension to the Semantic Web.
Though each individual chapter is clear, the book is a disconnected mixture of topics. The summary section at the end of each chapter consolidates the main discussion points, though in a simplistic manner. The first section lacks an overview of the relevance of ontologies in bioinformatics. There is no coherent definition of ontology and how it is different from information representation methods such as data models, database schemas and XML DTDs. The book puts forth a data model-centric perspective on the topic. The last chapter of the section is confusing because it starts out describing bio-ontologies but ends with a discussion on databases. In the second section, the chapters on using ontologies are a mixture of bioinformatics topics and an introduction to Perl, which is distracting. The concluding chapter on building ontologies is more relevant to the book. The final section on reasoning addresses a topic that is only tangentially relevant to ontologies.
The book would be most useful as a supplementary course textbook for an introductory bioinformatics course that covers ontologies and their use for data management. The ease with which each chapter reads will be appreciated by undergraduate students. Instructors can best utilize the book by selecting particular chapters that address the topics they are covering in the course, rather than using the book as a primary text. The book can be useful for introductory readings as well as a reference for students who wish to get a summary of the particular topics addressed by individual chapters.
Stanford Medical Informatics Stanford University, USA
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||