Overview of the SAMPL6 pKa challenge: evaluating small molecule microscopic and macroscopic pKa predictions.

The prediction of acid dissociation constants (pKa) is a prerequisite for predicting many other properties of a small molecule, such as its protein-ligand binding affinity, distribution coefficient (log D), membrane permeability, and solubility. The prediction of each of these properties requires knowledge of the relevant protonation states and solution free energy penalties of each state. The SAMPL6 pKa Challenge was the first time that a separate challenge was conducted for evaluating pKa predictions as part of the Statistical Assessment of Modeling of Proteins and Ligands (SAMPL) exercises. This challenge was motivated by significant inaccuracies observed in prior physical property prediction challenges, such as the SAMPL5 log D Challenge, caused by protonation state and pKa prediction issues. The goal of the pKa challenge was to assess the performance of contemporary pKa prediction methods for drug-like molecules. The challenge set was composed of 24 small molecules that resembled fragments of kinase inhibitors, a number of which were multiprotic. Eleven research groups contributed blind predictions for a total of 37 pKa distinct prediction methods. In addition to blinded submissions, four widely used pKa prediction methods were included in the analysis as reference methods. Collecting both microscopic and macroscopic pKa predictions allowed in-depth evaluation of pKa prediction performance. This article highlights deficiencies of typical pKa prediction evaluation approaches when the distinction between microscopic and macroscopic pKas is ignored; in particular, we suggest more stringent evaluation criteria for microscopic and macroscopic pKa predictions guided by the available experimental data. Top-performing submissions for macroscopic pKa predictions achieved RMSE of 0.7-1.0 pKa units and included both quantum chemical and empirical approaches, where the total number of extra or missing macroscopic pKas predicted by these submissions were fewer than 8 for 24 molecules. A large number of submissions had RMSE spanning 1-3 pKa units. Molecules with sulfur-containing heterocycles or iodo and bromo groups were less accurately predicted on average considering all methods evaluated. For a subset of molecules, we utilized experimentally-determined microstates based on NMR to evaluate the dominant tautomer predictions for each macroscopic state. Prediction of dominant tautomers was a major source of error for microscopic pKa predictions, especially errors in charged tautomers. The degree of inaccuracy in pKa predictions observed in this challenge is detrimental to the protein-ligand binding affinity predictions due to errors in dominant protonation state predictions and the calculation of free energy corrections for multiple protonation states. Underestimation of ligand pKa by 1 unit can lead to errors in binding free energy errors up to 1.2 kcal/mol. The SAMPL6 pKa Challenge demonstrated the need for improving pKa prediction methods for drug-like molecules, especially for challenging moieties and multiprotic molecules.

Standard state free energies, not pKas, are ideal for describing small molecule protonation and tautomeric states.

The pKa is the standard measure used to describe the aqueous proton affinity of a compound, indicating the proton concentration (pH) at which two protonation states (e.g. A- and AH) have equal free energy. However, compounds can have additional protonation states (e.g. AH2+), and may assume multiple tautomeric forms, with the protons in different positions (microstates). Macroscopic pKas give the pH where the molecule changes its total number of protons, while microscopic pKas identify the tautomeric states involved. As tautomers have the same number of protons, the free energy difference between them and their relative probability is pH independent so there is no pKa connecting them. The question arises: What is the best way to describe protonation equilibria of a complex molecule in any pH range? Knowing the number of protons and the relative free energy of all microstates at a single pH, ∆G°, provides all the information needed to determine the free energy, and thus the probability of each microstate at each pH. Microstate probabilities as a function of pH generate titration curves that highlight the low energy, observable microstates, which can then be compared with experiment. A network description connecting microstates as nodes makes it straightforward to test thermodynamic consistency of microstate free energies. The utility of this analysis is illustrated by a description of one molecule from the SAMPL6 Blind pKa Prediction Challenge. Analysis of microstate ∆G°s also makes a more compact way to archive and compare the pH dependent behavior of compounds with multiple protonatable sites.

pKa measurements for the SAMPL6 prediction challenge for a set of kinase inhibitor-like fragments.

Determining the net charge and protonation states populated by a small molecule in an environment of interest or the cost of altering those protonation states upon transfer to another environment is a prerequisite for predicting its physicochemical and pharmaceutical properties. The environment of interest can be aqueous, an organic solvent, a protein binding site, or a lipid bilayer. Predicting the protonation state of a small molecule is essential to predicting its interactions with biological macromolecules using computational models. Incorrectly modeling the dominant protonation state, shifts in dominant protonation state, or the population of significant mixtures of protonation states can lead to large modeling errors that degrade the accuracy of physical modeling. Low accuracy hinders the use of physical modeling approaches for molecular design. For small molecules, the acid dissociation constant (pKa) is the primary quantity needed to determine the ionic states populated by a molecule in an aqueous solution at a given pH. As a part of SAMPL6 community challenge, we organized a blind pKa prediction component to assess the accuracy with which contemporary pKa prediction methods can predict this quantity, with the ultimate aim of assessing the expected impact on modeling errors this would induce. While a multitude of approaches for predicting pKa values currently exist, predicting the pKas of drug-like molecules can be difficult due to challenging properties such as multiple titratable sites, heterocycles, and tautomerization. For this challenge, we focused on set of 24 small molecules selected to resemble selective kinase inhibitors-an important class of therapeutics replete with titratable moieties. Using a Sirius T3 instrument that performs automated acid-base titrations, we used UV absorbance-based pKa measurements to construct a high-quality experimental reference dataset of macroscopic pKas for the evaluation of computational pKa prediction methodologies that was utilized in the SAMPL6 pKa challenge. For several compounds in which the microscopic protonation states associated with macroscopic pKas were ambiguous, we performed follow-up NMR experiments to disambiguate the microstates involved in the transition. This dataset provides a useful standard benchmark dataset for the evaluation of pKa prediction methodologies on kinase inhibitor-like compounds.

Bayesian analysis of isothermal titration calorimetry for binding thermodynamics.

Isothermal titration calorimetry (ITC) is the only technique able to determine both the enthalpy and entropy of noncovalent association in a single experiment. The standard data analysis method based on nonlinear regression, however, provides unrealistically small uncertainty estimates due to its neglect of dominant sources of error. Here, we present a Bayesian framework for sampling from the posterior distribution of all thermodynamic parameters and other quantities of interest from one or more ITC experiments, allowing uncertainties and correlations to be quantitatively assessed. For a series of ITC measurements on metal:chelator and protein:ligand systems, the Bayesian approach yields uncertainties which represent the variability from experiment to experiment more accurately than the standard data analysis. In some datasets, the median enthalpy of binding is shifted by as much as 1.5 kcal/mol. A Python implementation suitable for analysis of data generated by MicroCal instruments (and adaptable to other calorimeters) is freely available online.

Acquired resistance to IDH inhibition through trans or cis dimer-interface mutations.

Somatic mutations in the isocitrate dehydrogenase 2 gene (IDH2) contribute to the pathogenesis of acute myeloid leukaemia (AML) through the production of the oncometabolite 2-hydroxyglutarate (2HG)1-8. Enasidenib (AG-221) is an allosteric inhibitor that binds to the IDH2 dimer interface and blocks the production of 2HG by IDH2 mutants9,10. In a phase I/II clinical trial, enasidenib inhibited the production of 2HG and induced clinical responses in relapsed or refractory IDH2-mutant AML11. Here we describe two patients with IDH2-mutant AML who had a clinical response to enasidenib followed by clinical resistance, disease progression, and a recurrent increase in circulating levels of 2HG. We show that therapeutic resistance is associated with the emergence of second-site IDH2 mutations in trans, such that the resistance mutations occurred in the IDH2 allele without the neomorphic R140Q mutation. The in trans mutations occurred at glutamine 316 (Q316E) and isoleucine 319 (I319M), which are at the interface where enasidenib binds to the IDH2 dimer. The expression of either of these mutant disease alleles alone did not induce the production of 2HG; however, the expression of the Q316E or I319M mutation together with the R140Q mutation in trans allowed 2HG production that was resistant to inhibition by enasidenib. Biochemical studies predicted that resistance to allosteric IDH inhibitors could also occur via IDH dimer-interface mutations in cis, which was confirmed in a patient with acquired resistance to the IDH1 inhibitor ivosidenib (AG-120). Our observations uncover a mechanism of acquired resistance to a targeted therapy and underscore the importance of 2HG production in the pathogenesis of IDH-mutant malignancies.

Biomolecular Simulations under Realistic Macroscopic Salt Conditions.

Biomolecular simulations are typically performed in an aqueous environment where the number of ions remains fixed for the duration of the simulation, generally with either a minimally neutralizing ion environment or a number of salt pairs intended to match the macroscopic salt concentration. In contrast, real biomolecules experience local ion environments where the salt concentration is dynamic and may differ from bulk. The degree of salt concentration variability and average deviation from the macroscopic concentration remains, as yet, unknown. Here, we describe the theory and implementation of a Monte Carlo osmostat that can be added to explicit solvent molecular dynamics or Monte Carlo simulations to sample from a semigrand canonical ensemble in which the number of salt pairs fluctuates dynamically during the simulation. The osmostat reproduces the correct equilibrium statistics for a simulation volume that can exchange ions with a large reservoir at a defined macroscopic salt concentration. To achieve useful Monte Carlo acceptance rates, the method makes use of nonequilibrium candidate Monte Carlo (NCMC) moves in which monovalent ions and water molecules are alchemically transmuted using short nonequilibrium trajectories, with a modified Metropolis-Hastings criterion ensuring correct equilibrium statistics for an ( Δµ, N, p, T) ensemble to achieve a ∼1046× boost in acceptance rates. We demonstrate how typical protein (DHFR and the tyrosine kinase Src) and nucleic acid (Drew-Dickerson B-DNA dodecamer) systems exhibit salt concentration distributions that significantly differ from fixed-salt bulk simulations and display fluctuations that are on the same order of magnitude as the average.

Binding Modes of Ligands Using Enhanced Sampling (BLUES): Rapid Decorrelation of Ligand Binding Modes via Nonequilibrium Candidate Monte Carlo.

Accurately predicting protein-ligand binding affinities and binding modes is a major goal in computational chemistry, but even the prediction of ligand binding modes in proteins poses major challenges. Here, we focus on solving the binding mode prediction problem for rigid fragments. That is, we focus on computing the dominant placement, conformation, and orientations of a relatively rigid, fragment-like ligand in a receptor, and the populations of the multiple binding modes which may be relevant. This problem is important in its own right, but is even more timely given the recent success of alchemical free energy calculations. Alchemical calculations are increasingly used to predict binding free energies of ligands to receptors. However, the accuracy of these calculations is dependent on proper sampling of the relevant ligand binding modes. Unfortunately, ligand binding modes may often be uncertain, hard to predict, and/or slow to interconvert on simulation time scales, so proper sampling with current techniques can require prohibitively long simulations. We need new methods which dramatically improve sampling of ligand binding modes. Here, we develop and apply a nonequilibrium candidate Monte Carlo (NCMC) method to improve sampling of ligand binding modes. In this technique, the ligand is rotated and subsequently allowed to relax in its new position through alchemical perturbation before accepting or rejecting the rotation and relaxation as a nonequilibrium Monte Carlo move. When applied to a T4 lysozyme model binding system, this NCMC method shows over 2 orders of magnitude improvement in binding mode sampling efficiency compared to a brute force molecular dynamics simulation. This is a first step toward applying this methodology to pharmaceutically relevant binding of fragments and, eventually, drug-like molecules. We are making this approach available via our new Binding modes of ligands using enhanced sampling (BLUES) package which is freely available on GitHub.

eTOX ALLIES: an automated pipeLine for linear interaction energy-based simulations.

BACKGROUND: Computational methods to predict binding affinities of small ligands toward relevant biological (off-)targets are helpful in prioritizing the screening and synthesis of new drug candidates, thereby speeding up the drug discovery process. However, use of ligand-based approaches can lead to erroneous predictions when structural and dynamic features of the target substantially affect ligand binding. Free energy methods for affinity computation can include steric and electrostatic protein-ligand interactions, solvent effects, and thermal fluctuations, but often they are computationally demanding and require a high level of supervision. As a result their application is typically limited to the screening of small sets of compounds by experts in molecular modeling. RESULTS: We have developed eTOX ALLIES, an open source framework that allows the automated prediction of ligand-binding free energies requiring the ligand structure as only input. eTOX ALLIES is based on the linear interaction energy approach, an efficient end-point free energy method derived from Free Energy Perturbation theory. Upon submission of a ligand or dataset of compounds, the tool performs the multiple steps required for binding free-energy prediction (docking, ligand topology creation, molecular dynamics simulations, data analysis), making use of external open source software where necessary. Moreover, functionalities are also available to enable and assist the creation and calibration of new models. In addition, a web graphical user interface has been developed to allow use of free-energy based models to users that are not an expert in molecular modeling. CONCLUSIONS: Because of the user-friendliness, efficiency and free-software licensing, eTOX ALLIES represents a novel extension of the toolbox for computational chemists, pharmaceutical scientists and toxicologists, who are interested in fast affinity predictions of small molecules toward biological (off-)targets for which protein flexibility, solvent and binding site interactions directly affect the strength of ligand-protein binding.

L-2-Hydroxyglutarate production arises from noncanonical enzyme function at acidic pH.

The metabolite 2-hydroxyglutarate (2HG) can be produced as either a D-R- or L-S- enantiomer, each of which inhibits α-ketoglutarate (αKG)-dependent enzymes involved in diverse biologic processes. Oncogenic mutations in isocitrate dehydrogenase (IDH) produce D-2HG, which causes a pathologic blockade in cell differentiation. On the other hand, oxygen limitation leads to accumulation of L-2HG, which can facilitate physiologic adaptation to hypoxic stress in both normal and malignant cells. Here we demonstrate that purified lactate dehydrogenase (LDH) and malate dehydrogenase (MDH) catalyze stereospecific production of L-2HG via 'promiscuous' reduction of the alternative substrate αKG. Acidic pH enhances production of L-2HG by promoting a protonated form of αKG that binds to a key residue in the substrate-binding pocket of LDHA. Acid-enhanced production of L-2HG leads to stabilization of hypoxia-inducible factor 1 alpha (HIF-1α) in normoxia. These findings offer insights into mechanisms whereby microenvironmental factors influence production of metabolites that alter cell fate and function.

Measuring experimental cyclohexane-water distribution coefficients for the SAMPL5 challenge.

Small molecule distribution coefficients between immiscible nonaqueuous and aqueous phases-such as cyclohexane and water-measure the degree to which small molecules prefer one phase over another at a given pH. As distribution coefficients capture both thermodynamic effects (the free energy of transfer between phases) and chemical effects (protonation state and tautomer effects in aqueous solution), they provide an exacting test of the thermodynamic and chemical accuracy of physical models without the long correlation times inherent to the prediction of more complex properties of relevance to drug discovery, such as protein-ligand binding affinities. For the SAMPL5 challenge, we carried out a blind prediction exercise in which participants were tasked with the prediction of distribution coefficients to assess its potential as a new route for the evaluation and systematic improvement of predictive physical models. These measurements are typically performed for octanol-water, but we opted to utilize cyclohexane for the nonpolar phase. Cyclohexane was suggested to avoid issues with the high water content and persistent heterogeneous structure of water-saturated octanol phases, since it has greatly reduced water content and a homogeneous liquid structure. Using a modified shake-flask LC-MS/MS protocol, we collected cyclohexane/water distribution coefficients for a set of 53 druglike compounds at pH 7.4. These measurements were used as the basis for the SAMPL5 Distribution Coefficient Challenge, where 18 research groups predicted these measurements before the experimental values reported here were released. In this work, we describe the experimental protocol we utilized for measurement of cyclohexane-water distribution coefficients, report the measured data, propose a new bootstrap-based data analysis procedure to incorporate multiple sources of experimental error, and provide insights to help guide future iterations of this valuable exercise in predictive modeling.

To Polarize or Not to Polarize? Charge-on-Spring versus KBFF Models for Water and Methanol Bulk and Vapor-Liquid Interfacial Mixtures.

Simulations of water and methanol mixtures using polarizable force fields (FFs) for methanol (COS/M and CPC) and water (COS/G2) were performed and compared to experiment and also to a nonpolarizable methanol (KBFF) model with SPC/E water in an effort to quantify the importance of explicit electronic polarization effects in bulk liquid mixtures and vapor-liquid interfaces. The bulk liquid mixture properties studied included the center of mass radial distribution functions, Kirkwood-Buff integrals (KBIs), volumetric properties, isothermal compressibility, enthalpy of mixing, dielectric constant, and diffusion coefficients. The vapor-liquid interface properties investigated included the relative surface probability distributions, surface tension, excess surface adsorption, preferred surface molecule orientations, and the surface dipole. None of the three FFs tested here was clearly superior for all of the properties examined. All the force fields typically reproduced the correct trends with composition for both the bulk and interfacial system properties; the differences between the force fields were primarily quantitative. The overall results suggest that the polarizable FFs are not, at the present stage of development, inherently better able to reproduce the studied bulk and interfacial properties-despite the added degree of explicit transferability that is, by definition, built into the polarizable models. Indeed, the specific parametrization of the FF appears to be just as important as the class of FF.

Toward Automated Benchmarking of Atomistic Force Fields: Neat Liquid Densities and Static Dielectric Constants from the ThermoML Data Archive.

Atomistic molecular simulations are a powerful way to make quantitative predictions, but the accuracy of these predictions depends entirely on the quality of the force field employed. Although experimental measurements of fundamental physical properties offer a straightforward approach for evaluating force field quality, the bulk of this information has been tied up in formats that are not machine-readable. Compiling benchmark data sets of physical properties from non-machine-readable sources requires substantial human effort and is prone to the accumulation of human errors, hindering the development of reproducible benchmarks of force-field accuracy. Here, we examine the feasibility of benchmarking atomistic force fields against the NIST ThermoML data archive of physicochemical measurements, which aggregates thousands of experimental measurements in a portable, machine-readable, self-annotating IUPAC-standard format. As a proof of concept, we present a detailed benchmark of the generalized Amber small-molecule force field (GAFF) using the AM1-BCC charge model against experimental measurements (specifically, bulk liquid densities and static dielectric constants at ambient pressure) automatically extracted from the archive and discuss the extent of data available for use in larger scale (or continuously performed) benchmarks. The results of even this limited initial benchmark highlight a general problem with fixed-charge force fields in the representation low-dielectric environments, such as those seen in binding cavities or biological membranes.

Hypoxia Induces Production of L-2-Hydroxyglutarate.

Somatic mutations in isocitrate dehydrogenase 1 or 2 (IDH1/2) contribute to the pathogenesis of cancer via production of the "oncometabolite" D-2-hydroxyglutarate (D-2HG). Elevated D-2HG can block differentiation of malignant cells by functioning as a competitive inhibitor of α-ketoglutarate (α-KG)-dependent enzymes, including Jumonji family histone lysine demethylases. 2HG is a chiral molecule that can exist in either the D-enantiomer or the L-enantiomer. Although cancer-associated IDH1/2 mutants produce D-2HG, biochemical studies have demonstrated that L-2HG also functions as a potent inhibitor of α-KG-dependent enzymes. Here we report that under conditions of oxygen limitation, mammalian cells selectively produce L-2HG via enzymatic reduction of α-KG. Hypoxia-induced L-2HG is not mediated by IDH1 or IDH2, but instead results from promiscuous substrate usage primarily by lactate dehydrogenase A (LDHA). During hypoxia, the resulting increase in L-2HG is necessary and sufficient for the induction of increased methylation of histone repressive marks, including histone 3 lysine 9 (H3K9me3).

QM/MM-Based Fitting of Atomic Polarizabilities for Use in Condensed-Phase Biomolecular Simulation.

Accounting for electronic polarization effects in biomolecular simulation (by using a polarizable force field) can increase the accuracy of simulation results. However, the use of gas-phase estimates of atomic polarizabilities αi usually leads to overpolarization in condensed-phase systems. In the current work, a combined QM/MM approach has been employed to obtain condensed-phase estimates of atomic polarizabilities for water and methanol (QM) solutes in the presence of (MM) solvents of different polarity. In a next step, the validity of the linear response and isotropy assumptions were evaluated based on the observed condensed-phase distributions of αi values. The observed anisotropy and low average values for the polarizability of methanol's carbon atom in polar solvents was explained in terms of strong solute-solvent interactions involving its adjacent hydroxyl group. Our QM/MM estimates for atomic polarizabilities were found to be close to values used in previously reported polarizable water and methanol models. Using our estimate for αO of methanol, a single set of polarizable force field parameters was obtained that is directly transferable between environments of different polarity.