Book Review: Quantitative In Silico Chromatography: Computational Modelling of Molecular Interactions

Chemistry is a beautiful science replete with chemical reactions that produce vividly colored solutions to megatons of clear liquids or white powders that are essential to our society.

“Better Living Thorough Chemistry” is much more than a great tag line.

In silico science, where the computer replaces the test tube, is intuitively attractive since it facilitates exploration of new ideas or compounds without as many experiments. For example, screening for new drugs involves synthesis of thousands of new compounds to fill out the library. If we had better models, we should be able to predict behavior and thus reduce the size of the chemical library, saving time, money, and waste.

Generally, in silico chemistry entails computer modeling of chemical processes to generate predicted structure‒activity relationships (SARs). Validation of the model involves comparing predictions of the SAR with experimental results. How close are we? The experiment is the judge. High regression (r) values show good agreement. Low r values mean one should refine or discard the model.

Over several decades, Prof. Toshihiko Hanai (Health Research Foundation, Kyoto, Japan) has published a series of papers that seek to quantitatively describe the kinetics and thermodynamics of the chromatographic process. His modus operandi is to create a homogeneous molecular model of a patch of stationary phase, then use computational programs to study the energetics of the probe compound (or ion) with the surface. The energetics includes van der Waals, hydrogen bonding, Lewis acid and base, and ionic interaction. Steric hindrance describes the spatial interaction of the stationary phase with the probe molecule. It seems to exclude entropic effects, which can be significant and even dominate the separation as in hydrophobic interaction chromatography (HIC).

In Quantitative In Silico Chromatography: Computational Modelling of Molecular Interactions,1 Hanai points out that advances in chromatographic technology have reduced separation time and improved selectivity and reproducibility. The separations are based on minute differences in the binding energy of the probe compounds with the stationary phase. This is the focus of his book. Chapters include adsorption, ion exchange, reversed phase, and affinity chromatography.

For example, in cation exchange, he modeled stationary phase with carboxylic acid groups, which is typical of many commercial weak ion exchangers (WCX). However, for anion exchange, the author’s model was a guanidine stationary phase. This surface chemistry is rare in commercial anion exchange column packings.

In the case of reversed-phase liquid chromatography (RPLC), there are more than 1000 C18 phases available. Most have different selectivity from others in this class. I did not see that Hanai’s in silico modeling provided insight into the SAR that accounts for these differences.

The quality of the models for various chromatographic modes varies considerably. In several cases, the r value is improved dramatically if outlier probes are removed from earlier, more inclusive, SAR sets. This may be expected, but one needs to consider whether the rejection is justified. One should also examine the winnowing process to see if it reveals an opportunity to improve the model. This is so obvious that I’m sure it was done. I’d like to see the reports.

In addition to preparative and analytical separation, chromatography has been used to probe molecular interactions, particularly in the life sciences. Affinity chromatography is an example where one can immobilize a drug receptor and probe a library for compounds that interact. Indeed, binding drugs to serum albumin is an example in the initial chapter and expanded on later in more detail.

Quantitative In Silico Chromatography will be a thought-provoking reference book for chromatographers and also chemists involved in association of molecules and ions. Prof. Hanai deserves special thanks for leading the development of molecular-scale modeling.

I expect that subsequent books on this topic will cite and build on this technology to improve Hanai’s useful correlations, thus refining the predictions of in silico SARs. Because of its high resolving power, chromatography will become an even more useful tool to study molecular interactions.

However, I expect that the detail-oriented reader will be frustrated by the poor image quality of the numerous models. For example, at least in my copy, I cannot easily see the analyte on the surface of the patch of stationary phase. When numbers representing the differences in charge are added, legibility declines rapidly. Higher-quality images would certainly have enhanced this work.

Reference

  1. Hanai, T. Quantitative In Silico Chromatography: Computational Modelling of Molecular Interactions. The Royal Society of Chemistry, Chromatography Monograph 19, 2014, ISBN978-1-84973-991-7.

Robert L. Stevenson, Ph.D., is Editor, American Laboratory/Labcompare; e-mail: [email protected].

Comments