Bernhard Gruber

Molecular libraries in liquid phase via UGI-MCR

Standard chemistry prescribes the conversion of one or two compounds into their products. In contrast, Eintopf (one-pot) multicomponent reactions (MCRs) involve at least three different compounds. One-pot MCRs are a useful tool in combinatorial chemistry: From a mixture of educts a large number of products can be simultaneously formed in liquid phase, called a soluble molecular library. The member compounds of such libraries are investigated simultaneously for desired properties, e.g. antibiotic activity.The main constraint is, that the underlying chemistry must not produce unknown side reactions and must lead to a broad spectrum of stable products with high yields.Isocyanide multicomponent chemistry allows the generation of soluble libraries of very different sizes, which are easy to screen for biological or pharmaceutical efficacy using the algorithms presented. Products can easily be enumerated and the kinetics of the isocyanide chemistry is simple to investigate.Combinatorial chemistry is capable of generating and optimizing leads faster and with fewer resources than by conventional means. Combinatorial chemistry based on isocyanide chemistry is by far the most important and most impressive technique in use today to reducing time and costs associated with lead generation and optimization during the drug discovery process. The simplicity of the reaction conditions involved means that the generation and screening of libraries can be automated.

Set-valued maps as a mathematical basis of computer assistance in stereochemistry

Previously, the theory of the chemical identity groups, an essentially nongeometric approach to stereochemistry, has been used for the representation and interpretation of permutational isomerizations and simple chemical reactions. Presently, the extension of this treatment to complex schemes of chemical reactions, such as the highly stereoselective syntheses of chiral a-ferrocenylalkylamines from (-)-menthone, is reported. Two computer programs for the application of the above theory and of the associated set-valued maps are described, a PROLOG program and “Chemld”, which is implemented i n Pascal.

MCR XVI. Mathematical support for combinatorial chemistry

The algebra of the s- and r-vectors is an adequate formal tool to describe chemical objects in an abstract way. Compounds as well as reactions are represented, including all constitutional and configurational aspects. The stereochemistry of simple organic molecules as well as those of metal-organic compounds may be described in a unique way. Ionic bonds, covalent bonds, aromatics, and electron-deficiency compounds can formally be described without loss of information. Even reaction types and the flow of electrons can be described by this algebra. The biggest benefit of this approach is its intrinsic group theoretical structure. This does not bother the chemist for its use but allows a computer to handle and structure huge amounts of chemical data. This is especially important for combinatorial chemistry, which deals with huge sets of chemically distinct molecules, the so-called molecular libraries.

MCR XII. Efficient Development of New Drugs by Online-Optimization of Molecular Libraries

The principles of combinatorial chemistry accelerate the development of new drugs enormously. Molecular libraries on the basis of multicomponent reactions (MCR) in liquid phase provide products of high diversity. Automated parallel synthesis and automatic analysis units (e.g. HPLC) provide the chance for optimizing the reaction conditions in an efficient way. Multi-parameter optimization by means of the genetic algorithm or other heuristics induce better yields at higher selectivity. The computer aided syntheses of molecular libraries under optimized reaction conditions with quality control results in an automaton, where the drug designer has the only but important task to choose the most useful starting compounds and thereby the molecular sub-space of a MCR.

MCR X. Important Aspects for Automating Preparative Chemistry

Usually, classical syntheses from n starting materials require sequences of at least n−1 preparation steps including separation and purification of the intermediates. A perfect alternative for the rapid synthesis of a large variety of chemical products are one-pot syntheses by multicomponent reactions (MCR)[1][2] based on the isocyanides. Between four and seven different types of reactants (educts) are mixed in a reaction vessel to form a product that contains at least one part of each educt. The educts and intermediates equilibrate and a stable product results, often with quantitative yields, in the final practically irreversible step involving the isocyanide. Reactions of this type are widely used for combinatorial chemistry. The minimisation of such syntheses and the computer-assisted handling of the results offer the chance of automating preparative chemistry.

MCR XI. Microreactors as Processors for Chemical Computers

Computer chemistry as well as the newer research in computational chemistry try to predict or estimate reaction pathways. To date this kind of syntheses planning has not succeeded, nor did approaches using rule-based heuristics like so-called expert systems. The reason for this is that reacting compounds are a highly parallel system. Molecules of the same chemical compound will react in different ways and/or at a different moment. Too many parameters are responsible for how chemical reactions proceed. Computers can not follow chemical reactions. Up to now, the only way to handle chemical problems is to use the chemistry itself. Usually synthetic chemists are interested in the main product exclusively. Naturally they do not care too much about side reactions and other strange things going on in their reaction vessels. But this complexity of chemical reactions can be useful for solving problems for which computers are far too slow.