Few experimental groups can afford the luxury of employing a theorist to work on their problems these days so they either model their data for themselves starting from scratch, or acquire a suite of modelling programs from a friend hoping they are applicable to their particular experiment. In either case, this book should be a great help, as it attempts to explain the basic underlying principles of modelling techniques for ensembles of small and moderate sized molecules. It should allow one to make a more informed choice between the various computational tools available. No prior knowledge of computer simulation is assumed, which makes this an ideal book for students and others just entering the field. However, there are several surprising omissions, for example the lack of any mention of the technique of reverse Monte Carlo (RMC) including the work of McGreevy, and any references to modelling societies such as CCP5 and MGMS (see below under useful URL addresses) where you can find others working on molecular modelling.
There are 4 sections, Basics, Ensembles, Phase Equilibria, and Advanced Techniques with a 5th section made up of 9 appendices, giving more detailed explanations of algorithms described briefly in the main sections. 'Basics' has clear explanations of Monte Carlo and Molecular Dynamics simulations. The algorithms are described in terms of a 'pseudo code' so that they can be implemented in the language of your choice. Each algorithm is surrounded by a box so that you can easily find them in the text; the 'pseudo code' is on the left of the box, an explanation of what it does on the right, in a different type face. In addition there are examples and 'Case studies' showing the application of the algorithm; these are also printed in a different font from the main text. Case studies of interest to crystallographers include a study of the absorption of methane in the zeolite silicalite, but most applications are to the properties of gases and liquids. The authors have coded all the case studies and made the FORTRAN source available on the Internet. (address below) An important topic, often ignored by those making computer simulations, is that of statistical errors, discussed in Appendix D.
The Appendices include a useful section on saving CPU time, including this diagram from page 368 Fig C.3, discussing the use of a cell list where the simulation cell is divided into cells of size c � c, where a particle only interacts with those particles in the same or neighboring cells (in 2D there are 9 cells; and in 3D 27cells).
The 8 page index has plenty of references to methods, but few to authors. For example if you want to find the reference to the text by Allen and Tildesley or Richard Catlow's book you have to search down the sequential list of 352 references in the bibliography.
The Suggested Reading section at the end of the Preface would have been improved by the inclusion of some more recent texts. For example, those looking for detailed computing information are referred to Allen and Tildesley's 'Computer Simulation of Liquids', an excellent book, but published in 1987, several years before the release of FORTRAN 90 with its matrix operations which drastically reduce the number of lines of code needed in this type of calculations. The suggested 'Modern Theoretical Chemistry' was published in 1977, long before the advent of the ubiquitous powerful workstations and microcomputers we have today. I got no feel for the type of computer which can best be used for these simulations. Do I need a supercomputer like a CRAY, or will my home micro suffice? Perhaps readers may get some guidance on the type of computer to choose by reading the report of the Daresbury Laboratory Evaluation Workshop reported in the CLRC news section elsewhere in this newsletter.
Conclusions A very good explanation of the basic principles of molecular simulation for ensembles of moderate sized molecules. There are sections on sampling polymer conformations but not on protein folding. You will have to look elsewhere for advice on detailed modern computing aspects.
Kate Crennell 30 October 1996
Click here to return to BCA homepage