The different strategies for designing GPCR and kinase targeted libraries.

In recent years the trend in combinatorial library design has shifted to include target class focusing along with diversity and drug-likeness criteria. In this manuscript we review the computational tools available for target class library design and highlight the areas where they have proven useful in our work. The protein kinase family is used to illustrated structure-based target class focused library design, and the G-protein coupled receptor (GPCR) family is used to illustrate ligand-based target class focused library design. Most of the tools discussed are those designed for libraries targeted to a single protein and are simply applied "brute-force" to a large number of targets within the family. The tools that have proven to be the most useful in our work are those that can extract trends from the computational data such as docking and clustering or data mining large amounts of structure activity or high throughput screening data. Finally, areas where improvements are needed in the computational tools available for target class focusing are highlighted. These areas include tools to extract the relevant patterns from all available information for a family of targets, tools to efficiently apply models for all targets in the family rather than just a small subset, mining tools to extract the relevant information from the computational absorption, distribution, metabolism, excretion and toxicity (ADMET) and targeting data, and tools to allow interactive exploration of the virtual space around a library to facilitate the selection of the library that best suits the needs of the design team.

High throughput docking for library design and library prioritization.

The prioritization of the screening of combinatorial libraries is an extremely important task for the rapid identification of tight binding ligands and ultimately pharmaceutical compounds. When structural information for the target is available, molecular docking is an approach that can be used for prioritization. Here, we present the initial validation of a new rapid approach to molecular docking developed for prioritizing combinatorial libraries. The algorithm is tested on 103 individual cases from the protein data bank and in nearly 90% of these cases docks the ligand to within 2.0 A of the observed binding mode. Because the mean CPU time is <5 s/mol, this approach can process hundreds of thousands of compounds per week. Furthermore, if a somewhat less thorough search is performed, the search time drops to 1 s/mol, thus allowing millions of compounds to be docked per week and tested for potential activity. Proteins 2001;43:113-124.

A database method for automated map interpretation in protein crystallography.

A significant portion of new protein structures contain folds that are related to those seen before. During the development of a computer program that can accurately position, in electron density maps, large protein domains with large structural deviations, it became apparent that the redundancy in protein folds could be used in a non trivial manner during a protein structure determination. As a result a computational procedure, Database Assisted Density Interpretation (DADI), was developed and tested to aid in the building of models in protein crystallography and to assist in interpreting electron density maps. The initial tests of the DADI procedure using a small database of protein domains are described. The philosophy is to first work with entire domains then with the secondary structure elements of these domains and finally with individual residues of the secondary structure elements via Monte Carlo, "chopping" and "clipping" procedures, respectively. The first test case was a traceable 3.2 A multiple isomorphous replacement with anomalous scattering (MIRAS) electron density map of a human topoisomerase I-DNA complex. The second test case uses poor electron density for the third domain of the diphtheria toxin repressor resulting from a molecular replacement solution with the first two domains. Despite the fact that a fairly small database was employed in these test cases, the DADI procedure was able to find a large portion of the protein backbone with very few errors. In the first case nearly 45% of the backbone and more than 80% of the secondary structure was placed automatically. In the second test case nearly 50% of the third domain was automatically detected. A particular encouraging result was that in both cases more than 75% of the beta sheet secondary structure was found automatically by the DADI procedure. Clearly, the procedures employed are promising avenues to exploit the current explosion of protein structures for the determination of future structures. Proteins 1999;36:526-541.

A rapid method for positioning small flexible molecules, nucleic acids, and large protein fragments in experimental electron density maps.

A Monte Carlo procedure, encoded in the program Blob, has been developed and tested for the purpose of positioning large molecular fragments or small flexible molecules in electron density maps. The search performed by the algorithm appears to be sufficiently thorough to accurately position a small flexible ligand in well-defined density while remaining sufficiently random to offer interesting alternate suggestions for density representing disordered binding modes of a ligand. Furthermore, the algorithm is shown to be efficient enough to accurately position large rigid molecular fragments. In the first of the test cases with large molecular fragments, Blob was surprisingly effective in positioning a poly-alanine model of a 53-residue domain in poor electron density resulting from molecular replacement with a partial model. At 3.0 A resolution the domain was positioned consistently within 0.2 A of its experimentally determined position. Even at 6.0 A resolution Blob could consistently position the domain to within 0.75 A of its actual position. A second set of tests with large molecular fragments revealed that Blob could correctly position large molecular fragments with quite significant deviations from the actual structure. In this test case, fragments ranging from a 170-residue protein domain with a 3.8 A rms deviation from the actual structure to a 22-base pair ideal B-form DNA duplex were positioned accurately in a 3.2 A electron density map derived from multiple isomorphous replacement methods. Even when decreasing the quality of the maps, from a figure of merit of 0.57 to as low as 0. 35, Blob could still effectively position the large protein domain and the DNA duplex. Since it is efficient, can handle large molecular fragments, and works in poor and low resolution maps, Blob could be a useful tool for interpreting electron density maps in de novo structure determinations and in molecular replacement studies. Proteins 1999;36:512-525.

The role of waters in docking strategies with incremental flexibility for carbohydrate derivatives: heat-labile enterotoxin, a multivalent test case.

Molecular docking studies of carbohydrate derivatives in protein binding sites are often challenging because of water-mediated interactions and the inherent flexibility of the many terminal hydroxyl groups. Using the recognition process between heat-labile enterotoxin from Escherichia coli and ganglioside GM1 as a paradigm, we developed a modeling protocol that includes incremental conformational flexibility of the ligand and predicted water interactions. The strategy employs a modified version of the Monte Carlo docking program AUTODOCK and water affinity potentials calculated with GRID. After calibration of the protocol on the basis of the known binding modes of galactose and lactose to the toxin, blind predictions were made for the binding modes of four galactose derivatives: lactulose, melibionic acid, thiodigalactoside, and m-nitrophenyl-alpha-galactoside. Subsequent crystal structure determinations have demonstrated that our docking strategy can predict the correct binding modes of carbohydrate derivatives within 1.0 A from experiment. In addition, it is shown that repeating the docking simulations in each of the seemingly identical binding sites of the multivalent toxin increases the chance of finding the correct binding mode.

An accurate numerical model for calculating the equilibration rate of a hanging-drop experiment.

A numerical model of the equilibration of a hanging-drop experiment has been developed and tested. To obtain accurate calculations with a given precipitant, the vapor pressure of water over water/precipitant solutions must be known for various concentrations of the precipitant. The calculations of the model are in excellent agreement with all available experimental data on hanging-drop equilibration when the necessary vapor pressures are known (ammonium sulfate and sodium chloride). By varying each of the relevant rate constants in the model, the rate-limiting step in the equilibration of a hanging drop is determined. This analysis clearly shows that the rate-limiting step is the diffusion of water vapor from the drop to the reservoir, which agrees with experimental findings. Since the diffusion of water vapor is the rate-limiting step, there is virtually no precipitant concentration gradient in the drop during equilibration. As a result, there is no gravity-induced convection owing to the equilibration. Thus, whereas gravity might have an effect during crystal growth, gravity does not affect the equilibration rate of a hanging-drop experiment to a significant extent, and the diffusion of water vapor will remain the rate-limiting step in the absence of gravity. Finally, the effects of several of the parameters, such as initial drop volume, drop-to-reservoir distance and temperature, are considered quantitatively. The equilibration rate was found to vary nearly linearly with drop volume. The equilibration rate decreases roughly by a factor of three as the temperature decreases from 293 to 276 K. This decrease in the equilibration rate is greater than would be expected when just considering the change in the diffusion coefficient of water vapor in air. This large dependence can, however, be attributed to the change in water-vapor pressure. Most surprisingly, a linear dependence on drop-to-reservoir distance is found, a result that agrees very well with experiment.