Knowledge Based Potentials: the Reverse Boltzmann Methodology, Virtual Screening and Molecular Weight Dependence

Ensemble docking to difficult targets in early-stage drug discovery: Methodology and application to fibroblast growth factor 23

Ensemble docking is now commonly used in early-stage in silico drug discovery and can be used to attack difficult problems such as finding lead compounds which can disrupt protein-protein interactions. We give an example of this methodology here, as applied to fibroblast growth factor 23 (FGF23), a protein hormone that is responsible for regulating phosphate homeostasis. The first small-molecule antagonists of FGF23 were recently discovered by combining ensemble docking with extensive experimental target validation data (Science Signaling, 9, 2016, ra113). Here, we provide a detailed account of how ensemble-based high-throughput virtual screening was used to identify the antagonist compounds discovered in reference (Science Signaling, 9, 2016, ra113). Moreover, we perform further calculations, redocking those antagonist compounds identified in reference (Science Signaling, 9, 2016, ra113) that performed well on drug-likeness filters, to predict possible binding regions. These predicted binding modes are rescored with the molecular mechanics Poisson-Boltzmann surface area (MM/PBSA) approach to calculate the most likely binding site. Our findings suggest that the antagonist compounds antagonize FGF23 through the disruption of protein-protein interactions between FGF23 and fibroblast growth factor receptor (FGFR).

Statistical potential for modeling and ranking of protein-ligand interactions

Applications in structural biology and medicinal chemistry require protein-ligand scoring functions for two distinct tasks: (i) ranking different poses of a small molecule in a protein binding site and (ii) ranking different small molecules by their complementarity to a protein site. Using probability theory, we developed two atomic distance-dependent statistical scoring functions: PoseScore was optimized for recognizing native binding geometries of ligands from other poses and RankScore was optimized for distinguishing ligands from nonbinding molecules. Both scores are based on a set of 8,885 crystallographic structures of protein-ligand complexes but differ in the values of three key parameters. Factors influencing the accuracy of scoring were investigated, including the maximal atomic distance and non-native ligand geometries used for scoring, as well as the use of protein models instead of crystallographic structures for training and testing the scoring function. For the test set of 19 targets, RankScore improved the ligand enrichment (logAUC) and early enrichment (EF(1)) scores computed by DOCK 3.6 for 13 and 14 targets, respectively. In addition, RankScore performed better at rescoring than each of seven other scoring functions tested. Accepting both the crystal structure and decoy geometries with all-atom root-mean-square errors of up to 2 Å from the crystal structure as correct binding poses, PoseScore gave the best score to a correct binding pose among 100 decoys for 88% of all cases in a benchmark set containing 100 protein-ligand complexes. PoseScore accuracy is comparable to that of DrugScore(CSD) and ITScore/SE and superior to 12 other tested scoring functions. Therefore, RankScore can facilitate ligand discovery, by ranking complexes of the target with different small molecules; PoseScore can be used for protein-ligand complex structure prediction, by ranking different conformations of a given protein-ligand pair. The statistical potentials are available through the Integrative Modeling Platform (IMP) software package (http://salilab.org/imp) and the LigScore Web server (http://salilab.org/ligscore/).

Informatics, machine learning and computational medicinal chemistry

This article reviews the use of informatics and computational chemistry methods in medicinal chemistry, with special consideration of how computational techniques can be adapted and extended to obtain more and higher-quality information. Special consideration is given to the computation of protein–ligand binding affinities, to the prediction of off-target bioactivities, bioactivity spectra and computational toxicology, and also to calculating absorption-, distribution-, metabolism- and excretion-relevant properties, such as solubility.

Potentials of mean force for protein structure prediction vindicated, formalized and generalized

Understanding protein structure is of crucial importance in science, medicine and biotechnology. For about two decades, knowledge-based potentials based on pairwise distances – so-called “potentials of mean force” (PMFs) – have been center stage in the prediction and design of protein structure and the simulation of protein folding. However, the validity, scope and limitations of these potentials are still vigorously debated and disputed, and the optimal choice of the reference state – a necessary component of these potentials – is an unsolved problem. PMFs are loosely justified by analogy to the reversible work theorem in statistical physics, or by a statistical argument based on a likelihood function. Both justifications are insightful but leave many questions unanswered. Here, we show for the first time that PMFs can be seen as approximations to quantities that do have a rigorous probabilistic justification: they naturally arise when probability distributions over different features of proteins need to be combined. We call these quantities “reference ratio distributions” deriving from the application of the “reference ratio method.” This new view is not only of theoretical relevance but leads to many insights that are of direct practical use: the reference state is uniquely defined and does not require external physical insights; the approach can be generalized beyond pairwise distances to arbitrary features of protein structure; and it becomes clear for which purposes the use of these quantities is justified. We illustrate these insights with two applications, involving the radius of gyration and hydrogen bonding. In the latter case, we also show how the reference ratio method can be iteratively applied to sculpt an energy funnel. Our results considerably increase the understanding and scope of energy functions derived from known biomolecular structures.

Scoring functions and their evaluation methods for protein-ligand docking: recent advances and future directions

The scoring function is one of the most important components in structure-based drug design. Despite considerable success, accurate and rapid prediction of protein-ligand interactions is still a challenge in molecular docking. In this perspective, we have reviewed three basic types of scoring functions (force-field, empirical, and knowledge-based) and the consensus scoring technique that are used for protein-ligand docking. The commonly-used assessment criteria and publicly available protein-ligand databases for performance evaluation of the scoring functions have also been presented and discussed. We end with a discussion of the challenges faced by existing scoring functions and possible future directions for developing improved scoring functions.

A machine learning approach to predicting protein-ligand binding affinity with applications to molecular docking

MOTIVATION

Accurately predicting the binding affinities of large sets of diverse protein-ligand complexes is an extremely challenging task. The scoring functions that attempt such computational prediction are essential for analysing the outputs of molecular docking, which in turn is an important technique for drug discovery, chemical biology and structural biology. Each scoring function assumes a predetermined theory-inspired functional form for the relationship between the variables that characterize the complex, which also include parameters fitted to experimental or simulation data and its predicted binding affinity. The inherent problem of this rigid approach is that it leads to poor predictivity for those complexes that do not conform to the modelling assumptions. Moreover, resampling strategies, such as cross-validation or bootstrapping, are still not systematically used to guard against the overfitting of calibration data in parameter estimation for scoring functions.

RESULTS

We propose a novel scoring function (RF-Score) that circumvents the need for problematic modelling assumptions via non-parametric machine learning. In particular, Random Forest was used to implicitly capture binding effects that are hard to model explicitly. RF-Score is compared with the state of the art on the demanding PDBbind benchmark. Results show that RF-Score is a very competitive scoring function. Importantly, RF-Score's performance was shown to improve dramatically with training set size and hence the future availability of more high-quality structural and interaction data is expected to lead to improved versions of RF-Score.

CONTACT

pedro.ballester@ebi.ac.uk; jbom@st-andrews.ac.uk

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Scoring functions and enrichment: a case study on Hsp90

Background

The need for fast and accurate scoring functions has been driven by the increased use of in silico virtual screening twinned with high-throughput screening as a method to rapidly identify potential candidates in the early stages of drug development. We examine the ability of some the most common scoring functions (GOLD, ChemScore, DOCK, PMF, BLEEP and Consensus) to discriminate correctly and efficiently between active and non-active compounds among a library of ~3,600 diverse decoy compounds in a virtual screening experiment against heat shock protein 90 (Hsp90).

Results

Firstly, we investigated two ranking methodologies, GOLD_rank and BestScore_rank. GOLD_rank is based on ranks generated using GOLD. The various scoring functions, GOLD, ChemScore, DOCK, PMF, BLEEP and Consensus, are applied to the pose ranked number one by GOLD for that ligand. BestScore_rank uses multiple poses for each ligand and independently chooses the best ranked pose of the ligand according to each different scoring function. Secondly, we considered the effect of introducing the Thr184 hydrogen bond tether to guide the docking process towards a particular solution, and its effect on enrichment. Thirdly, we considered normalisation to account for the known bias of scoring functions to select larger molecules. All the scoring functions gave fairly similar enrichments, with the exception of PMF which was consistently the poorest performer. In most cases, GOLD was marginally the best performing individual function; the Consensus score usually performed similarly to the best single scoring function. Our best results were obtained using the Thr184 tether in combination with the BestScore_rank protocol and normalisation for molecular weight. For that particular combination, DOCK was the best individual function; DOCK recovered 90% of the actives in the top 10% of the ranked list; Consensus similarly recovered 89% of the actives in its top 10%.

Conclusion

Overall, we demonstrate the validity of virtual screening as a method for identifying new leads from a pool of ligands with similar physicochemical properties and we believe that the outcome of this study provides useful insight into the setting up of a suitable docking and scoring protocol, resulting in enrichment of 'target active' compounds.