*Acta
Cryst.* (1994). B**50**,
479-480

**Pp. xvi + 271.
Weinheim: VCH Verlagsgesellschaft, 1992
Price $75.00. ISBN
1-56081-619-8**

This book is the third in a series intended to provide a guide to the rapidly developing field of computer-aided research in chemistry. The book contains four chapters written by experts in the field. It covers optimization methods, the prediction of oligopeptide structures, molecular modelling using NMR data and methods for the evaluation of chemical toxicity. A fifth chapter provides a compendium of molecular modelling software. The book is intended to provide a tutorial to novices, and a review for experts, of the theoretical background, the implementation methods and the advantages and disadvantages of popular computational chemistry approaches.

A knowledge of optimization methods and their limitations is necessary for any structural chemist who plans to use computational methods. These topics are covered in chapter 1 (71 pp., 142 references) by Tamar Schlick of New York University. He covers essential aspects of the subject, beginning with mathematical preliminaries and moving to search techniques and local and global methods of large-scale optimization, with an explanation of local descent methods and several varieties of minimization methods. This exposition is followed by examples of the performance of various minimization routines. The chapter is comprehensive and, in principle, should provide the education needed by users. However, the explanations are couched in mathematical symbolism rather than in words so that, while the chapter will be useful to the expert seeking a review of methods and implementation strategies, the novice who wishes to understand the basic function of these methods may not be particularly well served.

Chapter 2 (69 pp., 251 references), by Harold Scheraga of Cornell University, is optimistically entitled `Predicting three-dimensional structures of oligopeptides'. The chapter presents a detailed overview of the computational approaches adopted by the author and his co-workers, with only a cursory and generally critical discussion of alternative approaches to computing such structures or to solving the multiple-minima problem. There are summaries of the theoretical basis for the potential energy calculations used to predict structure, the build-up method of generating an oligopeptide chain and comparisons of computations with experimental observations. The presentation of the theory behind different potential functions and the methods available for treating hydration and entropy effects is clear, precise and useful.

Chapter 3 (49 pp., 121 references), by A. E. Torda and W. F. van Gunsteren of the ETH Zentrum, Zurich, presents the computational aspects of modelling a molecular structure based on nuclear magnetic resonance (NMR) data. The focus is on the determination of protein structure. The authors present the mathematical models for the experimental data and follow with an exposition of methods for refinement and minimization of the molecular model. They successfully bridge the gap between the novice reader and the expert, providing tuition for the former and a rich presentation of the state of the art of minimization and model development for the latter. The chapter provides a clear assessment of the errors and biases of the method and is instructive to the beginning user of molecular dynamics approaches. The presentation is clear and educational, and meets the objectives of the series.

Chapter 4 (49 pp., 227 references), by D. V. F. Lewis of the University of Surrey, presents computational approaches developed for the prediction of the toxicity of newly developed, untested chemicals. It is a good tutorial for the novice. A clear presentation of toxicology is provided and followed with excellent brief introductions to three computational approaches for toxicological study. Lewis clearly defines the parameters used in the traditional quantitative structure-activity relationship (QSAR) methods of Hansch, and explains pattern-recognition methods and knowledge-based systems. These methods are demonstrated and further explained by consideration of the P450 cytochromes and their role in toxic activation or detoxification of chemicals. Through this presentation, the author discusses several computer programs and their specific strengths. The chapter ends with a list of sources for these programs.

The editors conclude the book with a compendium of software, providing both a synopsis of and access information for 29 PC-based programs and 49 minicomputer or workstation-based programs for molecular modelling, quantum chemistry calculations, molecular graphics and molecular databases. This compendium is useful, but available in a much less expensive format from the Quantum Chemistry Program Exchange.

Overall, the book is elegantly presented and carefully edited. Two of the four chapters meet the stated aims of the series and provide a clear tutorial for novice readers. The book would be a useful addition to a research library; however, the lack of successful tutorial approaches for all four presentations makes it less desirable for a personal collection.

**Penelope W.
Codding**

*
Departments of Chemistry
and Pharmacology and
Therapeutics
University of
Calgary
Calgary
Alberta
Canada
T2N
1N4
*

**Copyright © 1997 International Union of Crystallography**