How does consensus scoring work for virtual library screening? An idealized computer experiment

Virtual screening to enrich hit lists from high-throughput screening: a case study on small-molecule inhibitors of angiogenin

"Hit lists" generated by high-throughput screening (HTS) typically contain a large percentage of false positives, making follow-up assays necessary to distinguish active from inactive substances. Here we present a method for improving the accuracy of HTS hit lists by computationally based virtual screening (VS) of the corresponding chemical libraries and selecting hits by HTS/VS consensus. This approach was applied in a case study on the target-enzyme angiogenin, a potent inducer of angiogenesis. In conjunction with HTS of the National Cancer Institute Diversity Set and ChemBridge DIVERSet E (approximately 18,000 compounds total), VS was performed with two flexible library docking/scoring methods, DockVision/Ludi and GOLD. Analysis of the results reveals that dramatic enrichment of the HTS hit rate can be achieved by selecting compounds in consensus with one or both of the VS functions. For example, HTS hits ranked in the top 2% by GOLD included 42% of the true hits, but only 8% of the false positives; this represents a sixfold enrichment over the HTS hit rate. Notably, the HTS/VS method was effective in selecting out inhibitors with midmicromolar dissociation constants typical of leads commonly obtained in primary screens.

Consideration of molecular weight during compound selection in virtual target-based database screening

Virtual database screening allows for millions of chemical compounds to be computationally selected based on structural complimentary to known inhibitors or to a target binding site on a biological macromolecule. Compound selection in virtual database screening when targeting a biological macromolecule is typically based on the interaction energy between the chemical compound and the target macromolecule. In the present study it is shown that this approach is biased toward the selection of high molecular weight compounds due to the contribution of the compound size to the energy score. To account for molecular weight during energy based screening, we propose normalization strategies based on the total number of heavy atoms in the chemical compounds being screened. This approach is computationally efficient and produces molecular weight distributions of selected compounds that can be selected to be (1) lower than that of the original database used in the virtual screening, which may be desirable for selection of leadlike compounds or (2) similar to that of the original database, which may be desirable for the selection of drug-like compounds. By eliminating the bias in target-based database screening toward higher molecular weight compounds it is anticipated that the proposed procedure will enhance the success rate of computer-aided drug design.

Structure-based virtual screening: an overview

Enormous advances in genomics have resulted in a large increase in the number of potential therapeutic targets that are available for investigation. This growth in potential targets has increased the demand for reliable target validation, as well as technologies that can identify rapidly several quality lead candidates. Virtual screening, and in particular receptor-based virtual screening, has emerged as a reliable, inexpensive method for identifying leads. Although still an evolving method, advances in computational techniques have enabled virtual screening to have a positive impact on the discovery process. Here, the current strengths and weaknesses of the technology are discussed, and emphasis is placed on aspects of the work-flow of a virtual screening campaign, from preparation through to post-screening analysis.

DrugScore meets CoMFA: adaptation of fields for molecular comparison (AFMoC) or how to tailor knowledge-based pair-potentials to a particular protein

The development of a new tailor-made scoring function to predict binding affinities of protein-ligand complexes is described. Knowledge-based pair-potentials are specifically adapted to a particular protein by considering additional ligand-based information. The formalism applied to derive the new function is similar to the well-known CoMFA approach, however, the fields used in the approach originate from the protein environment (and not from the aligned ligands as in CoMFA, thus, a "reverse" CoMFA (= AFMoC) named Adaptation of Fields for Molecular Comparison is performed). A regular-spaced grid is placed into the binding site and knowledge-based pair-potentials between protein atoms and ligand atom probes are mapped onto the grid intersections resulting in "potential fields". By multiplying distance-dependent atom-type properties of actual ligands docked into the binding site with the neighboring grid values, "interaction fields" are produced from the original "potential fields". In a PLS analysis, these atom-type specific interaction fields are correlated to the actual binding affinities of the embedded ligands, resulting in individual weighting factors for each field value. As in CoMFA, the results of the analysis can be interpreted in graphical terms by contribution maps, and binding affinities of novel ligands are predicted by applying the derived 3D QSAR equation. The scope of the new method is demonstrated using thermolysin and glycogen phosphorylase b as test examples. Impressive improvements of the predictive power for affinity prediction can be achieved compared to the application of the original knowledge-based potentials by considering a sample set of only 15 known training ligands. Thus, with growing information about the drug target studied, the new method allows one to move gradually from generally valid to protein-specifically adapted pair-potentials, depending on the amount of training information available and its degree of structural diversity. In addition, convincing predictive power is also achieved for ligand poses generated by automatic docking tools.

ConsDock: A new program for the consensus analysis of protein-ligand interactions

Protein-based virtual screening of chemical libraries is a powerful technique for identifying new molecules that may interact with a macromolecular target of interest. Because of docking and scoring limitations, it is more difficult to apply as a lead optimization method because it requires that the docking/scoring tool is able to propose as few solutions as possible and all of them with a very good accuracy for both the protein-bound orientation and the conformation of the ligand. In the present study, we present a consensus docking approach (ConsDock) that takes advantage of three widely used docking tools (Dock, FlexX, and Gold). The consensus analysis of all possible poses generated by several docking tools is performed sequentially in four steps: (i) hierarchical clustering of all poses generated by a docking tool into families represented by a leader; (ii) definition of all consensus pairs from leaders generated by different docking programs; (iii) clustering of consensus pairs into classes, represented by a mean structure; and (iv) ranking the different means starting from the most populated class of consensus pairs. When applied to a test set of 100 protein-ligand complexes from the Protein Data Bank, ConsDock significantly outperforms single docking with respect to the docking accuracy of the top-ranked pose. In 60% of the cases investigated here, ConsDock was able to rank as top solution a pose within 2 A RMSD of the X-ray structure. It can be applied as a postprocessing filter to either single- or multiple-docking programs to prioritize three-dimensional guided lead optimization from the most likely docking solution.

Further development and validation of empirical scoring functions for structure-based binding affinity prediction

New empirical scoring functions have been developed to estimate the binding affinity of a given protein-ligand complex with known three-dimensional structure. These scoring functions include terms accounting for van der Waals interaction, hydrogen bonding, deformation penalty, and hydrophobic effect. A special feature is that three different algorithms have been implemented to calculate the hydrophobic effect term, which results in three parallel scoring functions. All three scoring functions are calibrated through multivariate regression analysis of a set of 200 protein-ligand complexes and they reproduce the binding free energies of the entire training set with standard deviations of 2.2 kcal/mol, 2.1 kcal/mol, and 2.0 kcal/mol, respectively. These three scoring functions are further combined into a consensus scoring function, X-CSCORE. When tested on an independent set of 30 protein-ligand complexes. X-CSCORE is able to predict their binding free energies with a standard deviation of 2.2 kcal/mol. The potential application of X-CSCORE to molecular docking is also investigated. Our results show that this consensus scoring function improves the docking accuracy considerably when compared to the conventional force field computation used for molecular docking.