W. Todd Wipke

Hydride as a leaving group in the reaction of pinacolborane with halides under ambient Grignard and Barbier conditions. One-pot synthesis of alkyl, aryl, heteroaryl, vinyl, and allyl pinacolboronic esters.

Grignard reagents (aliphatic, aromatic, heteroaromatic, vinyl, or allylic) react with 1 equiv of 4,4,5,5-tetramethyl-1,3,2-dioxaborolane (pinacolborane, PinBH) at ambient temperature in tetrahydrofuran (THF) to afford the corresponding pinacolboronates. The initially formed dialkoxy alkylborohydride intermediate quickly eliminates hydridomagnesium bromide (HMgBr) and affords the product boronic ester in very good yield. Hydridomagnesium bromide (HMgBr) in turn disproportionates to a 1:1 mixture of magnesium hydride (MgH(2)) and magnesium bromide (MgBr(2)) on addition of pentane to the reaction mixture. DFT calculations (Gaussian09) at the B3LYP/6-31G(d) level of theory show that disproportionation of HMgBr to MgH(2) and MgBr(2) is viable in the coordinating ethereal solvents. This reaction also can be carried out under Barbier conditions, where the neat PinBH is added to the flask prior to the in situ formation of Grignard reagent from the corresponding organic halide and magnesium metal. Pinacolboronic ester synthesis under Barbier conditions does not give Wurtz coupling side products from reactive halides, such as benzylic and allylic halides. The reaction between PinBH and various Grignard reagents is an efficient, mild, and general method for the synthesis of pinacolboronates.

QSD quadratic shape descriptors. 2. Molecular docking using quadratic shape descriptors (QSDock)

We present a new shape‐based polynomial time algorithm for the rapid docking of rigid ligands into their macromolecular receptors. The method exploits molecular surface complementarity existing between a putative ligand and its receptor protein. Molecular shapes are represented by using a new shape descriptor that is based on local quadratic approximations to the molecular surface. The quadratic shape descriptor is capable of representing a plethora of molecular shapes and is not limited to describing convex or concave regions of molecular surface. A single pair of complementary descriptors is sufficient for computing the transformation matrix that positions a ligand into the receptor site. We demonstrate the capabilities of our algorithm by successfully reproducing the crystallographically determined orientation for a test set of 20 ligand‐protein complexes. Proteins 2000;38:79–94.

Simulation and evaluation of chemical synthesis—SECS: an application of artificial intelligence techniques

The problem of designing chemical syntheses of complex organic compounds is a challenging domain for application of artificial intelligence techniques. SECS is an interactive program to assist a chemist in heuristically searching and evaluating the space of good synthetic pathways. The chemist-computer team, linked through computer graphics, develops synthetic plans using a logic-centered backward analysis from the target structure. The reaction knowledge base, written in the ALCHEM language, is separate from the program and control strategies. Performance is demonstrated on the insect pheromone grandisol.

Artificial intelligence in organic synthesis. SST: starting material selection strategies. An application of superstructure search.

A program for the interactive selection of potential starting materials given a desired target molecule is described. The program uses hierarchical search to rapidly select candidates from a large starting material library and contains a function to evaluate the appropriateness of the functionality of the starting material. The user can specify restrictions on features such as the number of atoms in the starting material, price limitations, chirality, and whether to use superstructure or substructure searching. Several examples of the results of the program are presented.

AIMB: Analogy and intelligence in model building. System description and performance characteristics

Abstract This paper describes a new approach to building molecular models using reasoning by analogy. We apply symbolic reasoning to a problem previously only approached numerically. AIMB mimics the behavior of expert chemists: the more knowledge AIMB has, the faster and more accurate the models AIMB builds. AIMB has rules of analogy in the program, but all knowledge of geometry is in the knowledge base of known 3D structures. The method described here generates minimum energy models without minimization and is significantly faster than the fastest numerical minimization methods. MolFiles of models produced by AIMB and of the corresponding crystal structures are included on disk 4.

Rapid subgraph search using parallelism.

A new parallel processing algorithm is reported for subgraph matching. Parallelism is achieved for the first time within the process of node-by-node matching of two individual graphs. A SIMULA program is described for simulating this parallel subgraph search algorithm. Simulation results from a series of chemical substructure search problems show an average utilization of 84% on a 25-processor machine and up to a 24-fold speed enhancement over a single processor. Potential applications include starting material selections for synthesis as well as general substructure search problems.

Database structure and searching in MACCS-3D

Abstract The MACCS-3D program was developed to provide a three-dimensional (3D) structural database system for structure, substructure, geometric, and data searching purposes. Applications of the program include discovery of new drugs and agrochemicals to fit known or postulated pharmacophores, and identification of polymers and commercial chemicals whose physical and chemical properties depend on 3D structural features. This paper presents details of the data structures and algorithms used for 3D structural representation, storage, and searching in MACCS-3D. These include algorithms for exact match, substructure, geometric, submodel, and combined 3D structure-data searching. We also describe mechanisms used for 3D data storage and geometric search keys. Integration of MACCS-3D with new chemical representation and search capabilities of the MACCS-II system are briefly described. Finally, new developments in MACCS-3D are described, in light of current and future trends in 3D searching.

Quadratic shape descriptors. 1. Rapid superposition of dissimilar molecules using geometrically invariant surface descriptors.

In this paper, we present a novel approach to shape-based molecular similarity searching. The method that we introduce is able to superimpose dissimilar molecules by using geometrically invariant molecular surface descriptors. The shape descriptors are calculated by least-squares fitting of a quadratic function to small sections of the molecular surface of a ligand. Invariant geometric properties of the approximated surface patch are then extracted from the fitted quadratic function. The extracted properties are used to quantify the shape and to obtain a canonical orientation for this section of surface. The superimposition algorithm uses these geometric invariants to recognize similar regions of surface shape existing on two molecules and to bring these regions (and consequently the molecules) into registration. Because these geometric descriptors are based upon local surface shape, the superimposing algorithm is insensitive to the connectivity and the relative sizes of the molecules being matched. The capabilities of our algorithm are demonstrated by superimposing dissimilar ligands known to inhibit the same enzyme system. In all cases examined the algorithm generates superpositions that are in agreement with crystallographic results. The algorithm is also applied to align the two different proteins on the basis of the shape of their active sites.

CLUSMOL: A system for the conceptual clustering of molecules☆

CLUSMOL is designed to conceptualize the chemists understanding of a group of molecules, by creating taxonomic trees, in which structurally similar molecules are placed in the close leaf nodes, while the intermediary nodes represent the maximal common substructures for the leaf structures under the node. The system is tested on some antibiotics and shown to be useful for lead structure recognition in drug design.

Automatic knowledge base building for the organic synthesis design program (SECS)

Abstract SECS (Simulation and Evaluation of Chemical Synthesis) is a retrosynthetic organic synthesis design program. Building a large knowledge base is required in order to solve important synthesis problems in industry. We developed a method that automatically builds a large ALCHEM transform library. Fifteen thousand reactions have been built. These were modified and reorganized into a form usable by SECS. At the same time, we added an additional strategy generation function to the SECS strategy module. We present our method and an example of a SECS analysis of a target molecule using our large ALCHEM knowledge base.