Thomas E. Exner

Empirical Scoring Functions for Advanced Protein−Ligand Docking with PLANTS

In this paper we present two empirical scoring functions, PLANTS(CHEMPLP) and PLANTS(PLP), designed for our docking algorithm PLANTS (Protein-Ligand ANT System), which is based on ant colony optimization (ACO). They are related, regarding their functional form, to parts of already published scoring functions and force fields. The parametrization procedure described here was able to identify several parameter settings showing an excellent performance for the task of pose prediction on two test sets comprising 298 complexes in total. Up to 87% of the complexes of the Astex diverse set and 77% of the CCDC/Astex clean listnc (noncovalently bound complexes of the clean list) could be reproduced with root-mean-square deviations of less than 2 A with respect to the experimentally determined structures. A comparison with the state-of-the-art docking tool GOLD clearly shows that this is, especially for the druglike Astex diverse set, an improvement in pose prediction performance. Additionally, optimized parameter settings for the search algorithm were identified, which can be used to balance pose prediction reliability and search speed.

An ant colony optimization approach to flexible protein-ligand docking

Abstract The prediction of the complex structure of a small ligand with a protein, the so-called protein–ligand docking problem, is a central part of the rational drug design process. For this purpose, we introduce the docking algorithm PLANTS (Protein–Ligand ANT System), which is based on ant colony optimization, one of the most successful swarm intelligence techniques. We study the effectiveness of PLANTS for several parameter settings and present a direct comparison of PLANTS’s performance to a state-of-the-art program called GOLD, which is based on a genetic algorithm and frequently used in the pharmaceutical industry for this task. Last but not least, we also show that PLANTS can make effective use of protein flexibility giving example results on cross-docking and virtual screening experiments for protein kinase A.

PLANTS: application of ant colony optimization to structure-based drug design

A central part of the rational drug development process is the prediction of the complex structure of a small ligand with a protein, the so-called protein-ligand docking problem, used in virtual screening of large databases and lead optimization. In the work presented here, we introduce a new docking algorithm called PLANTS (Protein-Ligand ANTSystem), which is based on ant colony optimization. An artificial ant colony is employed to find a minimum energy conformation of the ligand in the protein’s binding site. We present the effectiveness of PLANTS for several parameter settings as well as a direct comparison to a state-of-the-art program called GOLD, which is based on a genetic algorithm. Last but not least, results for a virtual screening on the protein target factor Xa are presented.

Ab initio quality properties for macromolecules using the ADMA approach

We describe new developments of an earlier linear scaling algorithm for ab initio quality macromolecular property calculations based on the adjustable density matrix assembler (ADMA) approach. In this approach, a large molecule is divided into fuzzy fragments, for which quantum chemical calculations can easily be done using moderate‐size “parent molecules” that contain all the local interactions within a selected distance. If greater accuracy is required, a larger distance is chosen. With the present extension of this approximation, properties of the large molecules, like the electron density, the electrostatic potential, dipole moments, partial charges, and the Hartree–Fock energy are calculated. The accuracy of the method is demonstrated with test cases of medium size by comparing the ADMA results with direct quantum chemical calculations.

Influence of Protonation, Tautomeric, and Stereoisomeric States on Protein−Ligand Docking Results

In this work, we present a systematical investigation of the influence of ligand protonation states, stereoisomers, and tautomers on results obtained with the two protein-ligand docking programs GOLD and PLANTS. These different states were generated with a fully automated tool, called SPORES (Structure PrOtonation and Recognition System). First, the most probable protonations, as defined by this rule based system, were compared to the ones stored in the well-known, manually revised CCDC/ASTEX data set. Then, to investigate the influence of the ligand protonation state on the docking results, different protonation states were created. Redocking and virtual screening experiments were conducted demonstrating that both docking programs have problems in identifying the correct protomer for each complex. Therefore, a preselection of plausible protomers or the improvement of the scoring functions concerning their ability to rank different molecules/states is needed. Additionally, ligand stereoisomers were tested for a subset of the CCDC/ASTEX set, showing similar problems regarding the ranking of these stereoisomers as the ranking of the protomers.

Toward the Quantum Chemical Calculation of NMR Chemical Shifts of Proteins : 2. Level of Theory, Basis Set, and Solvents Model Dependence

It has been demonstrated that the fragmentation scheme of our adjustable density matrix assembler (ADMA) approach for the quantum chemical calculations of very large systems is well-suited to calculate NMR chemical shifts of proteins [ Frank et al. Proteins2011, 79, 2189-2202 ]. The systematic investigation performed here on the influences of the level of theory, basis set size, inclusion or exclusion of an implicit solvent model, and the use of partial charges to describe additional parts of the macromolecule on the accuracy of NMR chemical shifts demonstrates that using a valence triple-ζ basis set leads to large improvement compared to the results given in the previous publication. Additionally, moving from the B3LYP to the mPW1PW91 density functional and including partial charges and implicit solvents gave the best results with mean absolute errors of 0.44 ppm for hydrogen atoms excluding H(N) atoms and between 1.53 and 3.44 ppm for carbon atoms depending on the size and also on the accuracy of the protein structure. Polar hydrogen and nitrogen atoms are more difficult to predict. For the first, explicit hydrogen bonds to the solvents need to be included and, for the latter, going beyond DFT to post-Hartree-Fock methods like MP2 is probably required. Even if empirical methods like SHIFTX+ show similar performance, our calculations give for the first time very reliable chemical shifts that can also be used for complexes of proteins with small-molecule ligands or DNA/RNA. Therefore, taking advantage of its ab initio nature, our approach opens new fields of application that would otherwise be largely inaccessible due to insufficient availability of data for empirical parametrization.

Pattern recognition strategies for molecular surfaces: III. Binding site prediction with a neural network

An algorithm for the identification of possible binding sites of biomolecules, which are represented as regions of the molecular surface, is introduced. The algorithm is based on the segmentation of the molecular surface into overlapping patches as described in the first article of this series. 1 The properties of these patches (calculated on the basis of physical and chemical properties) are used for the analysis of the molecular surfaces of 7821 proteins and protein complexes. Special attention is drawn to known protein binding sites. A binding site identification algorithm is realized on the basis of the calculated data using a neural network strategy. The neural network is able to classify surface patches as protein–protein, protein–DNA, protein–ligand, or nonbinding sites. To show the capability of the algorithm, results of the surface analysis and the predictions are presented and discussed with representative examples.

pKa based protonation states and microspecies for protein–ligand docking

In this paper we present our reworked approach to generate ligand protonation states with our structure preparation tool SPORES (Structure PrOtonation and REcognition System). SPORES can be used for the preprocessing of proteins and protein–ligand complexes as e.g. taken from the Protein Data Bank as well as for the setup of 3D ligand databases. It automatically assigns atom and bond types, generates different protonation, tautomeric states as well as different stereoisomers. In the revised version, pKa calculations with the ChemAxon software MARVIN are used either to determine the likeliness of a combinatorial generated protonation state or to determine the titrable atoms used in the combinatorial approach. Additionally, the MARVIN software is used to predict microspecies distributions of ligand molecules. Docking studies were performed with our recently introduced program PLANTS (Protein–Ligand ANT System) on all protomers resulting from the three different selection methods for the well established CCDC/ASTEX clean data set demonstrating the usefulness of especially the latter approach.

Toward the quantum chemical calculation of nuclear magnetic resonance chemical shifts of proteins

Despite the many protein structures solved successfully by nuclear magnetic resonance (NMR) spectroscopy, quality control of NMR structures is still by far not as well established and standardized as in crystallography. Therefore, there is still the need for new, independent, and unbiased evaluation tools to identify problematic parts and in the best case also to give guidelines that how to fix them. We present here, quantum chemical calculations of NMR chemical shifts for many proteins based on our fragment‐based quantum chemical method: the adjustable density matrix assembler (ADMA). These results show that 13C chemical shifts of reasonable accuracy can be obtained that can already provide a powerful measure for the structure validation. 1H and even more 15N chemical shifts deviate more strongly from experiment due to the insufficient treatment of solvent effects and conformational averaging. Proteins 2011;

Discovery of Two Classes of Potent Glycomimetic Inhibitors of Pseudomonas aeruginosa LecB with Distinct Binding Modes

The treatment of infections due to the opportunistic pathogen Pseudomonas aeruginosa is often difficult, as a consequence of bacterial biofilm formation. Such a protective environment shields the bacterium from host defense and antibiotic treatment and secures its survival. One crucial factor for maintenance of the biofilm architecture is the carbohydrate-binding lectin LecB. Here, we report the identification of potent mannose-based LecB inhibitors from a screening of four series of mannosides in a novel competitive binding assay for LecB. Cinnamide and sulfonamide derivatives are inhibitors of bacterial adhesion with up to a 20-fold increase in affinity to LecB compared to the natural ligand methyl mannoside. Because many lectins of the host require terminal saccharides (e.g., fucosides), such capped structures as reported here may offer a beneficial selectivity profile for the pathogenic lectin. Both classes of compounds show distinct binding modes at the protein, offering the advantage of a simultaneous development of two new lead structures as anti-pseudomonadal drugs with an anti-virulence mode of action.