Computational Analysis of Biochemical Systems : A Practical Guide for Biochemists and Molecular Biologists

Rating:
Summary: Biochemical Systems Approach is Fundamentally Flawed
Review: This book details the Biochemical Systems Theory approach to modeling and simulating biochemical systems. Example applications are provided, along with the PLAS (Power Law Analysis and Simulation) software for convenient implementation of the author's approach. The book focuses on the so-called S-system approach to modeling biochemical transformations, an approach introduced by Voit, Savegeau, and co-workers.

The book fails to address important alternative approaches to biochemical kinetics. Namely, the mechanistic approach to modeling enzymatic reactions, which the majority of biochemical kinetic theory is based, is ignored. This omission is particularly disastrous because S-systems fail to satisfy mass balance and to realistically represent reaction thermodynamics. The approaches outlined in this book should be abandoned because: (1) The S-systems in this book (and elsewhere) do not conserve mass. The stoichiometry of reaction networks is lost in their approach and mass balance, a fundamental cornerstone concept is ignored. (2) The abandonment of stoichiometry results in a theory which ignores thermodynamics altogether. The resulting unphysical approach is a poor foundation on which to build integrated systems models.

Rating:
Summary: fundamentals
Review: This is a well-executed entry level ODE solver the authors have included into a system of pull down GUI that make the process much easier. Standard time series analysis using ODE including phase plane, flux, and eigenvalue analysis are all incorporated under one beaker sort of speak. The most disturbing feature is the lack of a print command in the pull down menu and the index. Fortunately the main text is read by MS Word and will print from there. The graphs must be printed out through Excel, which is good given its greater flexibility. After the construction of a few well-directed Excel macros this problem disappeared. Direct insertion into Matlab through their Excel link allows Plas to be used as a DDE extension in parallel programming. The Insert feature (table) in Mathcad provides the same avenue. However Plas is so easy to use I find myself going to it rather than one of the bigger programs. The GUI interface table for the steady states and eigenvalue are just too easy. Obviously the more one knows about this type of analysis the more one appreciates the product however by the same token if one does know more they are probably using more sophisticated techniques. The really good part is the book however. An exceptionally well written and clear text, it would be good as introductory material in any computer (using a computer for dry lab scientific work) modeling environment without the freebe program. The methodology is suitable for Biophysics as well as Economics because the basic Plas method is a universal solution set whose methods deals with systems (ODEs). The authors refer to them as the S system, but they are simple cascaded linear ordinary differential equations. The reader was able to model a biophysics retinal adaptation model using 5 equations and 2 feedback loops involving amacrine, cone and in about 15 min. to check figures against one done in Matlab and Mathcad. The same models can be built using 1st and 2nd order difference equations with iterated range variables in the two math programs. This program uses somewhat different ODE algorithms (the linearized part of Taylor's expansion) which is similar in numerical values to your basic Runga-Kutta. While the book material and program are written more for the linear analysis done in today's science stuff (referred to as 'the steady state') the subject material of main concern of this reader is the non-linear events and deterministic chaos where all the really interesting stuff is located. The built in graphing mode of Plas allows for 2 and 3D phase projection and eigenvalue analysis which is the basic required inputs for bifurcation and limit cycle analysis which makes this book and program a best buy. If bigger stuff (example, stability by Hankel reduction or system state conversion to zpk optimization) is on your list the "readability of its files" doesn't cause obsolescence

Rating:
Summary: An important bridge between biology and computing
Review: This work is an excellent, thorough presentation of a methodology for modeling and simulating biochemical systems. The book intelligently avoids the twin perils of being too superficial (website listings masquerading as bioinformatics text books) and being too academic (theoretical models that cannot be applied by biologists without a second PhD.)

Instead, a single modeling methodology called Biochemical Systems Theory is completely presented, from basics to advanced case studies of biological applications. The author begins by presenting a graphical method of modeling biochemical systems, then logically proceeds to explain a useful and powerful mathematical approach to simulating the behavior of modeled systems by using Power-Law methods. Finally, the modeling and simulation techniques are applied in several case studies, showing the reader what is most important: what can be learned about biology by the intelligent application of a computer simulation. Installable, working software is distributed along with the book, providing readers with instant access to the tools and models used in the case studies.

As any reasonable, experienced scientist or practitioner can attest to, every modeling approach to a problem has strengths and limitations. The author, a well known and established pioneer, demonstrates high integrity and candor in acquainting the reader with all of the challenges, as well as the power, of this particular modeling methodology.

What makes this book so important is its presentation of a complete methodology for modeling and problem solving. Readers who are truly interested in innovative yet proven methods for understanding the dynamics of real world biochemical systems will find this work extremely useful and relevant. Those who are looking for a broad survey of current methods should probably consult the literature for an updated review of the latest developments.

And finally, those few individuals who are quixotically pining away for "the perfect model" will probably need to relocate to the parallel universe where such impossibilities may actually exist. ;) In this universe, the best we can do is apply useful models and methodologies to gain real insight into the hard questions--with help from useful and well written books like this one.