Briefings in Bioinformatics Advance Access originally published online on December 19, 2006
Briefings in Bioinformatics 2007 8(2):134-135; doi:10.1093/bib/bbl018
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Book Reviews |
The Ten Most Wanted Solutions in Protein Bioinformatics.
Anna Tramontano
The Ten Most Wanted Solutions in Protein Bioinformatics.
Anna Tramontano
Chapman & Hall/CRC;
ISBN: 1584884916; 216pp.; 2005
From the size and title of the book, Anna Tramontano has taken on a difficult task: describe ten challenges in protein bioinformatics in less than 200 pages. The author is not content to cover ten topics in any restricted sense; each problem (chapter) addresses broad categories, such as, sequence alignment, function prediction or membrane proteins. Taken as a set, the ten problems just about span the entire protein bioinformatics space. Given the brisk pace of research all of these areas, discussing the plethora of approaches could easily fill 200 pages for each problem. The author has chosen to describe the classic solutions with some mention of the ‘state of the art’ addressed primarily to an audience with training in biology or biochemistry and with some familiarity with protein biophysics. Unlike many short texts on bioinformatics, which describe the development of bioinformatics databases and are somewhat sketchy on algorithms, Dr. Tramontano briskly covers the mathematical foundation of the most widely used methods. It is a difficult material to present concisely, but her approach is well-conceived and her prose is clear and nicely reinforced with appropriate figures.
The book's content is apropos for the field: from protein sequence alignment to structure and function prediction, with due consideration for the exciting developments in our understanding of membrane proteins, the author spans familiar and reliable methods in the field (sequence alignment, protein structure prediction) to problems, such as, protein design and prediction of protein–protein interactions where we may lack adequate means to fully test new methodology. Methods discussed in each chapter focus on those with the longest standing and broadest applicability. The discussion is conceptual, with some historical insight into the development of the methodology. Coverage of each topic is somewhat uneven: some chapters are devoted entirely to bioinformatics methodology (sequence alignment); other chapters interweave descriptions of experimental methods with computational approaches (protein–protein interactions). Given the space constraints, which preclude exhaustive description and evaluation of current methodology, the author sensibly refers the audience to efforts within each community, such as, the well-known Critical Assessment of Techniques for Protein Structure Prediction (CASP) or less familiar Critical Assessment of Predicted Interactions (CAPRI). However, a summary of some of the evaluative criteria from CASP and CAPRI would be beneficial. Unfortunately, not every community within the realm of protein bioinformatics run assessment events and, therefore, even a short discussion devoted to evaluation of methodology (when such criteria exist) for each chapter would have been appreciated. This is the book's weakest point. The suggested readings reflect the content of the book. There is a reasonably short list of readings; some are classic articles or textbooks, other readings are devoted to methods, which are exemplars for their category.
This book is a good introduction for someone who is familiar with bioinformatics in a more applied sense and who now wishes to understand the theoretical foundations or for someone who understands one aspect of bioinformatics and is interested in a quick introduction to other topics in the field. It is not the most appropriate first book to recommend to someone completely unfamiliar with bioinformatics, unless they are from a related field, such as, computational structural biology or are otherwise comfortable with algorithms. Unfortunately, significant biological background is assumed, otherwise this would be an excellent volume for computer scientists. The introduction serves very well as a quick test to determine if the reader has sufficient conceptual understanding of biological and biophysical concepts to read the book comfortably. The lack of any exercises renders this book unsuitable for use as textbook but it would serve as useful supplemental reading for an introductory class in bioinformatics at the senior undergraduate/graduate level where the majority of the students enter with training in biology. For anyone with an active interest in bioinformatics, ‘Ten Problems’ is an appropriate place to start reading, but should be paired with another textbook, which would develop appropriate skills to either use current methodologies or to develop one's own.
At this point, I must interject a small protest. First, the layout of the text and graphics, while adequate for the pre-Internet age, is inferior to even basic web technology. Graphics interspersed with the text are monochrome, color figures are relegated to a central section; the disruption is noticeable. Secondly, the illustration of the protein structures is generally poor. Unfortunately, Figure 1 draws your attention to this problem immediately. As my eyes were confronted with a mash of poorly differentiated grey tones and my hand reached towards a hyperlink to a color version that didn’t exist, I wondered if anyone had considered the necessary conversion of what was clearly a color original for black-and-white presentation. Upon redirection of the hand towards the more appropriate task of flipping to the section with the color graphics, I was disappointed again with a very small graphic, with hues and saturation values best dealt with by my closing my eyes. Most of the depictions of proteins are not optimized to convey the 3D-nature of the structure, which can be achieved either by stereoscopic renderings (for the cross-eyed) or by introducing differences in color saturation from near to far. Finally, after examination of the graphics, I cannot detect any awareness of the need to convert graphics developed for presentation on a computer monitor to graphics for publication. I am relieved to report the problems are confined to depiction of protein structures and the rest of the diagrams are clear.
Stanford Medical Informatics
Stanford University
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||