Evaluation and Characterization of Trk Kinase Inhibitors for the Treatment of Pain: Reliable Binding Affinity Predictions from Theory and Computation

Machine Learning Methods as a Cost-Effective Alternative to Physics-Based Binding Free Energy Calculations

The rank ordering of ligands remains one of the most attractive challenges in drug discovery. While physics-based in silico binding affinity methods dominate the field, they still have problems, which largely revolve around forcefield accuracy and sampling. Recent advances in machine learning have gained traction for protein-ligand binding affinity predictions in early drug discovery programs. In this article, we perform retrospective binding free energy evaluations for 172 compounds from our internal collection spread over four different protein targets and five congeneric ligand series. We compared multiple state-of-the-art free energy methods ranging from physics-based methods with different levels of complexity and conformational sampling to state-of-the-art machine-learning-based methods that were available to us. Overall, we found that physics-based methods behaved particularly well when the ligand perturbations were made in the solvation region, and they did not perform as well when accounting for large conformational changes in protein active sites. On the other end, machine-learning-based methods offer a good cost-effective alternative for binding free energy calculations, but the accuracy of their predictions is highly dependent on the experimental data available for training the model.

Accelerating COVID-19 Drug Discovery with High-Performance Computing

The recent COVID-19 pandemic has served as a timely reminder that the existing drug discovery is a laborious, expensive, and slow process. Never has there been such global demand for a therapeutic treatment to be identified as a matter of such urgency. Unfortunately, this is a scenario likely to repeat itself in future, so it is of interest to explore ways in which to accelerate drug discovery at pandemic speed. Computational methods naturally lend themselves to this because they can be performed rapidly if sufficient computational resources are available. Recently, high-performance computing (HPC) technologies have led to remarkable achievements in computational drug discovery and yielded a series of new platforms, algorithms, and workflows. The application of artificial intelligence (AI) and machine learning (ML) approaches is also a promising and relatively new avenue to revolutionize the drug design process and therefore reduce costs. In this review, I describe how molecular dynamics simulations (MD) were successfully integrated with ML and adapted to HPC to form a powerful tool to study inhibitors for four of the COVID-19 target proteins. The emphasis of this review is on the strategy that was used with an explanation of each of the steps in the accelerated drug discovery workflow. For specific technical details, the reader is directed to the relevant research publications.

The performance of ensemble-based free energy protocols in computing binding affinities to ROS1 kinase

Optimization of binding affinities for compounds to their target protein is a primary objective in drug discovery. Herein we report on a collaborative study that evaluates a set of compounds binding to ROS1 kinase. We use ESMACS (enhanced sampling of molecular dynamics with approximation of continuum solvent) and TIES (thermodynamic integration with enhanced sampling) protocols to rank the binding free energies. The predicted binding free energies from ESMACS simulations show good correlations with experimental data for subsets of the compounds. Consistent binding free energy differences are generated for TIES and ESMACS. Although an unexplained overestimation exists, we obtain excellent statistical rankings across the set of compounds from the TIES protocol, with a Pearson correlation coefficient of 0.90 between calculated and experimental activities.

Alchemical Free Energy Estimators and Molecular Dynamics Engines: Accuracy, Precision, and Reproducibility

The binding free energy between a ligand and its target protein is an essential quantity to know at all stages of the drug discovery pipeline. Assessing this value computationally can offer insight into where efforts should be focused in the pursuit of effective therapeutics to treat a myriad of diseases. In this work, we examine the computation of alchemical relative binding free energies with an eye for assessing reproducibility across popular molecular dynamics packages and free energy estimators. The focus of this work is on 54 ligand transformations from a diverse set of protein targets: MCL1, PTP1B, TYK2, CDK2, and thrombin. These targets are studied with three popular molecular dynamics packages: OpenMM, NAMD2, and NAMD3 alpha. Trajectories collected with these packages are used to compare relative binding free energies calculated with thermodynamic integration and free energy perturbation methods. The resulting binding free energies show good agreement between molecular dynamics packages with an average mean unsigned error between them of 0.50 kcal/mol. The correlation between packages is very good, with the lowest Spearman's, Pearson's and Kendall's tau correlation coefficients being 0.92, 0.91, and 0.76, respectively. Agreement between thermodynamic integration and free energy perturbation is shown to be very good when using ensemble averaging.

Ensemble Simulations and Experimental Free Energy Distributions: Evaluation and Characterization of Isoxazole Amides as SMYD3 Inhibitors

Optimization of binding affinities for ligands to their target protein is a primary objective in rational drug discovery. Herein, we report on a collaborative study that evaluates various compounds designed to bind to the SET and MYND domain-containing protein 3 (SMYD3). SMYD3 is a histone methyltransferase and plays an important role in transcriptional regulation in cell proliferation, cell cycle, and human carcinogenesis. Experimental measurements using the scintillation proximity assay show that the distributions of binding free energies from a large number of independent measurements exhibit non-normal properties. We use ESMACS (enhanced sampling of molecular dynamics with approximation of continuum solvent) and TIES (thermodynamic integration with enhanced sampling) protocols to predict the binding free energies and to provide a detailed chemical insight into the nature of ligand-protein binding. Our results show that the 1-trajectory ESMACS protocol works well for the set of ligands studied here. Although one unexplained outlier exists, we obtain excellent statistical ranking across the set of compounds from the ESMACS protocol and good agreement between calculations and experiments for the relative binding free energies from the TIES protocol. ESMACS and TIES are again found to be powerful protocols for the accurate comparison of the binding free energies.

Large Scale Study of Ligand-Protein Relative Binding Free Energy Calculations: Actionable Predictions from Statistically Robust Protocols

The accurate and reliable prediction of protein-ligand binding affinities can play a central role in the drug discovery process as well as in personalized medicine. Of considerable importance during lead optimization are the alchemical free energy methods that furnish an estimation of relative binding free energies (RBFE) of similar molecules. Recent advances in these methods have increased their speed, accuracy, and precision. This is evident from the increasing number of retrospective as well as prospective studies employing them. However, such methods still have limited applicability in real-world scenarios due to a number of important yet unresolved issues. Here, we report the findings from a large data set comprising over 500 ligand transformations spanning over 300 ligands binding to a diverse set of 14 different protein targets which furnish statistically robust results on the accuracy, precision, and reproducibility of RBFE calculations. We use ensemble-based methods which are the only way to provide reliable uncertainty quantification given that the underlying molecular dynamics is chaotic. These are implemented using TIES (Thermodynamic Integration with Enhanced Sampling). Results achieve chemical accuracy in all cases. Ensemble simulations also furnish information on the statistical distributions of the free energy calculations which exhibit non-normal behavior. We find that the "enhanced sampling" method known as replica exchange with solute tempering degrades RBFE predictions. We also report definitively on numerous associated alchemical factors including the choice of ligand charge method, flexibility in ligand structure, and the size of the alchemical region including the number of atoms involved in transforming one ligand into another. Our findings provide a key set of recommendations that should be adopted for the reliable application of RBFE methods.

Thermodynamic and structural insights into the repurposing of drugs that bind to SARS-CoV-2 main protease

Although researchers have been working tirelessly since the COVID-19 outbreak, so far only three drugs - remdesivir, ronapreve and molnupiravir - have been approved for use in some countries which directly target the SARS-CoV-2 virus. Given the slow pace and substantial costs of new drug discovery and development, together with the urgency of the matter, repurposing of existing drugs for the ongoing disease is an attractive proposition. In a recent study, a high-throughput X-ray crystallographic screen was performed for a selection of drugs which have been approved or are in clinical trials. Thirty-seven compounds have been identified from drug libraries all of which bind to the SARS-CoV-2 main protease (3CL^pro). In the current study, we use molecular dynamics simulation and an ensemble-based free energy approach, namely, enhanced sampling of molecular dynamics with approximation of continuum solvent (ESMACS), to investigate a subset of the aforementioned compounds. The drugs studied here are highly diverse, interacting with different binding sites and/or subsites of 3CL^pro. The predicted free energies are compared with experimental results wherever they are available and they are found to be in excellent agreement. Our study also provides detailed energetic insights into the nature of the associated drug-protein binding, in turn shedding light on the design and discovery of potential drugs.

Pandemic drugs at pandemic speed: infrastructure for accelerating COVID-19 drug discovery with hybrid machine learning- and physics-based simulations on high-performance computers

The race to meet the challenges of the global pandemic has served as a reminder that the existing drug discovery process is expensive, inefficient and slow. There is a major bottleneck screening the vast number of potential small molecules to shortlist lead compounds for antiviral drug development. New opportunities to accelerate drug discovery lie at the interface between machine learning methods, in this case, developed for linear accelerators, and physics-based methods. The two in silico methods, each have their own advantages and limitations which, interestingly, complement each other. Here, we present an innovative infrastructural development that combines both approaches to accelerate drug discovery. The scale of the potential resulting workflow is such that it is dependent on supercomputing to achieve extremely high throughput. We have demonstrated the viability of this workflow for the study of inhibitors for four COVID-19 target proteins and our ability to perform the required large-scale calculations to identify lead antiviral compounds through repurposing on a variety of supercomputers.

The effect of protein mutations on drug binding suggests ensuing personalised drug selection

The advent of personalised medicine promises a deeper understanding of mechanisms and therefore therapies. However, the connection between genomic sequences and clinical treatments is often unclear. We studied 50 breast cancer patients belonging to a population-cohort in the state of Qatar. From Sanger sequencing, we identified several new deleterious mutations in the estrogen receptor 1 gene (ESR1). The effect of these mutations on drug treatment in the protein target encoded by ESR1, namely the estrogen receptor, was achieved via rapid and accurate protein-ligand binding affinity interaction studies which were performed for the selected drugs and the natural ligand estrogen. Four nonsynonymous mutations in the ligand-binding domain were subjected to molecular dynamics simulation using absolute and relative binding free energy methods, leading to the ranking of the efficacy of six selected drugs for patients with the mutations. Our study shows that a personalised clinical decision system can be created by integrating an individual patient's genomic data at the molecular level within a computational pipeline which ranks the efficacy of binding of particular drugs to variant proteins.

Uncertainty quantification in classical molecular dynamics

Molecular dynamics simulation is now a widespread approach for understanding complex systems on the atomistic scale. It finds applications from physics and chemistry to engineering, life and medical science. In the last decade, the approach has begun to advance from being a computer-based means of rationalizing experimental observations to producing apparently credible predictions for a number of real-world applications within industrial sectors such as advanced materials and drug discovery. However, key aspects concerning the reproducibility of the method have not kept pace with the speed of its uptake in the scientific community. Here, we present a discussion of uncertainty quantification for molecular dynamics simulation designed to endow the method with better error estimates that will enable it to be used to report actionable results. The approach adopted is a standard one in the field of uncertainty quantification, namely using ensemble methods, in which a sufficiently large number of replicas are run concurrently, from which reliable statistics can be extracted. Indeed, because molecular dynamics is intrinsically chaotic, the need to use ensemble methods is fundamental and holds regardless of the duration of the simulations performed. We discuss the approach and illustrate it in a range of applications from materials science to ligand-protein binding free energy estimation. This article is part of the theme issue 'Reliability and reproducibility in computational science: implementing verification, validation and uncertainty quantification in silico'.

Recent Advances in Pain Management: Relevant Protein Kinases and Their Inhibitors

The purpose of this review is to underline the protein kinases that have been established, either in fundamental approach or clinical trials, as potential biological targets in pain management. Protein kinases are presented according to their group in the human kinome: TK (Trk, RET, EGFR, JAK, VEGFR, SFK, BCR-Abl), CMGC (p38 MAPK, MEK, ERK, JNK, ASK1, CDK, CLK2, DYRK1A, GSK3, CK2), AGC (PKA, PKB, PKC, PKMζ, PKG, ROCK), CAMK, CK1 and atypical/other protein kinases (IKK, mTOR). Examples of small molecule inhibitors of these biological targets, demonstrating an analgesic effect, are described. Altogether, this review demonstrates the fundamental role that protein kinase inhibitors could play in the development of new pain treatments.

Rapid, accurate, precise and reproducible ligand-protein binding free energy prediction

A central quantity of interest in molecular biology and medicine is the free energy of binding of a molecule to a target biomacromolecule. Until recently, the accurate prediction of binding affinity had been widely regarded as out of reach of theoretical methods owing to the lack of reproducibility of the available methods, not to mention their complexity, computational cost and time-consuming procedures. The lack of reproducibility stems primarily from the chaotic nature of classical molecular dynamics (MD) and the associated extreme sensitivity of trajectories to their initial conditions. Here, we review computational approaches for both relative and absolute binding free energy calculations, and illustrate their application to a diverse set of ligands bound to a range of proteins with immediate relevance in a number of medical domains. We focus on ensemble-based methods which are essential in order to compute statistically robust results, including two we have recently developed, namely thermodynamic integration with enhanced sampling and enhanced sampling of MD with an approximation of continuum solvent. Together, these form a set of rapid, accurate, precise and reproducible free energy methods. They can be used in real-world problems such as hit-to-lead and lead optimization stages in drug discovery, and in personalized medicine. These applications show that individual binding affinities equipped with uncertainty quantification may be computed in a few hours on a massive scale given access to suitable high-end computing resources and workflow automation. A high level of accuracy can be achieved using these approaches.

Hit-to-lead and lead optimization binding free energy calculations for G protein-coupled receptors

We apply the hit-to-lead ESMACS (enhanced sampling of molecular dynamics with approximation of continuum solvent) and lead-optimization TIES (thermodynamic integration with enhanced sampling) methods to compute the binding free energies of a series of ligands at the A1 and A2A adenosine receptors, members of a subclass of the GPCR (G protein-coupled receptor) superfamily. Our predicted binding free energies, calculated using ESMACS, show a good correlation with previously reported experimental values of the ligands studied. Relative binding free energies, calculated using TIES, accurately predict experimentally determined values within a mean absolute error of approximately 1 kcal mol-1. Our methodology may be applied widely within the GPCR superfamily and to other small molecule-receptor protein systems.

Large-scale binding affinity calculations on commodity compute clouds

In recent years, it has become possible to calculate binding affinities of compounds bound to proteins via rapid, accurate, precise and reproducible free energy calculations. This is imperative in drug discovery as well as personalized medicine. This approach is based on molecular dynamics (MD) simulations and draws on sequence and structural information of the protein and compound concerned. Free energies are determined by ensemble averages of many MD replicas, each of which requires hundreds of cores and/or GPU accelerators, which are now available on commodity cloud computing platforms; there are also requirements for initial model building and subsequent data analysis stages. To automate the process, we have developed a workflow known as the binding affinity calculator. In this paper, we focus on the software infrastructure and interfaces that we have developed to automate the overall workflow and execute it on commodity cloud platforms, in order to reliably predict their binding affinities on time scales relevant to the domains of application, and illustrate its application to two free energy methods.

Advances in the calculation of binding free energies

In recent years, calculations of binding affinities from molecular simulations seem to have matured significantly. While the number of applications of such methods in drug design and biotechnology increases, the number of truly new methodological developments decreases. This review provides an overview of the current status of the field as reflected in recent publications. The focus is on the challenges that remain when using endstate, alchemical and pathway methods. For endstate methods this is the calculation of entropic contributions. For alchemical methods there are unsolved problems associated with the solvation of the active site, sampling slow degrees of freedom and when modifying the net charge. For pathway methods achieving sufficient sampling remains challenging. New trends are also highlighted, including the use of pathway methods for the quantification of protein-protein interactions.

Accuracy and Precision of Alchemical Relative Free Energy Predictions with and without Replica-Exchange

A systematic and statistically robust protocol is applied for the evaluation of free energy calculations with and without replica-exchange. The protocol is based on ensemble averaging to generate accurate assessments of the uncertainties in the predictions. Comparison is made between FEP+ and TIES-free energy perturbation and thermodynamic integration with enhanced sampling-the latter with and without the so-called "enhanced sampling" based on replica-exchange protocols. Standard TIES performs best for a reference set of targets and compounds; no benefits accrue from replica-exchange methods. Evaluation of FEP+ and TIES with REST-replica-exchange with solute tempering-reveals a systematic and significant underestimation of free energy differences in FEP+, which becomes increasingly large for long duration simulations, is confirmed by extensive analysis of previous publications, and raises a number of questions pertaining to the accuracy of the predictions with the REST technique not hitherto discussed.

Application of the ESMACS Binding Free Energy Protocol to a Multi-Binding Site Lactate Dehydogenase A Ligand Dataset

Over the past two decades, the use of fragment-based lead generation has become a common, mature approach to identify tractable starting points in chemical space for the drug discovery process. This approach naturally involves the study of the binding properties of highly heterogeneous ligands. Such datasets challenge computational techniques to provide comparable binding free energy estimates from different binding modes. The performance of a range of statistically robust ensemble-based binding free energy calculation protocols, called ESMACS (enhanced sampling of molecular dynamics with approximation of continuum solvent), is evaluated. Ligands designed to target two binding pockets in the lactate dehydogenase, a target protein, which vary in size, charge, and binding mode, are studied. When compared to experimental results, excellent statistical rankings are obtained across this highly diverse set of ligands. In addition, three approaches to account for entropic contributions are investigated: 1) normal mode analysis, 2) weighted solvent accessible surface area (WSAS), and 3) variational entropy. Normal mode analysis and WSAS correlate strongly with each other-although the latter is computationally far cheaper-but do not improve rankings. Variational entropy corrects exaggerated discrimination of ligands bound in different pockets but creates three outliers which reduce the quality of the overall ranking.

DeltaDelta neural networks for lead optimization of small molecule potency

The capability to rank different potential drug molecules against a protein target for potency has always been a fundamental challenge in computational chemistry due to its importance in drug design. While several simulation-based methodologies exist, they are hard to use prospectively and thus predicting potency in lead optimization campaigns remains an open challenge. Here we present the first machine learning approach specifically tailored for ranking congeneric series based on deep 3D-convolutional neural networks. Furthermore we prove its effectiveness by blindly testing it on datasets provided by Janssen, Pfizer and Biogen totalling over 3246 ligands and 13 targets as well as several well-known openly available sets, representing one the largest evaluations ever performed. We also performed online learning simulations of lead optimization using the approach in a predictive manner obtaining significant advantage over experimental choice. We believe that the evaluation performed in this study is strong evidence of the usefulness of a modern deep learning model in lead optimization pipelines against more expensive simulation-based alternatives.

Application of ESMACS binding free energy protocols to diverse datasets: Bromodomain-containing protein 4

As the application of computational methods in drug discovery pipelines becomes more widespread it is increasingly important to understand how reproducible their results are and how sensitive they are to choices made in simulation setup and analysis. Here we use ensemble simulation protocols, termed ESMACS (enhanced sampling of molecular dynamics with approximation of continuum solvent), to investigate the sensitivity of the popular molecular mechanics Poisson-Boltzmann surface area (MMPBSA) methodology. Using the bromodomain-containing protein 4 (BRD4) system bound to a diverse set of ligands as our target, we show that robust rankings can be produced only through combining ensemble sampling with multiple trajectories and enhanced solvation via an explicit ligand hydration shell.

Targeting ERK1/2 protein-serine/threonine kinases in human cancers

ERK1 and ERK2 are key protein kinases that contribute to the Ras-Raf-MEK-ERK MAP kinase signalling module. This pathway participates in the control of numerous processes including apoptosis, cell proliferation, the immune response, nervous system function, and RNA synthesis and processing. MEK1/2 activate human ERK1/2 by first catalyzing the phosphorylation of Y204/187 and then T202/185, both residues of which occur within the activation segment. The phosphorylation of both residues is required for enzyme activation. The only Raf substrates are MEK1/2 and the only MEK1/2 substrates are ERK1/2. In contrast, ERK1/2 catalyze the phosphorylation of many cytoplasmic and nuclear substrates including transcription factors and regulatory molecules. The linear MAP kinase pathway branches extensively at the ERK1/2 node. ERK1/2 are proline-directed kinases that preferentially catalyze the phosphorylation of substrates containing a PxS/TP sequence. The dephosphorylation and inactivation of ERK1/2 is catalyzed by dual specificity phosphatases, protein-tyrosine specific phosphatases, and protein-serine/threonine phosphatases. The combined functions of kinases and phosphatases make the overall process reversible. To provide an idea of the complexities involved in these reactions, somatic cell cycle progression involves the strict timing of more than 32,000 phosphorylation and dephosphorylation events as determined by mass spectrometry. The MAP kinase cascade is perhaps the most important oncogenic driver of human cancers and the blockade of this signalling module by targeted inhibitors is an important anti-tumor strategy. Although numerous cancers are driven by MAP kinase pathway activation, thus far the only orally effective approved drugs that target this signaling module are used for the treatment of BRAF-mutant melanomas. The best treatments include the combination of B-Raf and MEK inhibitors (dabrafenib and trametinib, encorafenib and binimetinib, vemurafenib and cobimetanib). However, resistance to these antagonists occurs within one year and additional treatment options are necessary. Owing to the large variety of malignancies that are driven by dysregulation of the MAP kinase pathway, additional tumor types should be amenable to MAP kinase pathway inhibitor therapy. In addition to new B-Raf and MEK inhibitors, the addition of ERK inhibitors should prove helpful. Ulixertinib, MK-8353, and GDC-0994 are orally effective, potent, and specific inhibitors of ERK1/2 that are in early clinical trials for the treatment of various advanced/metastatic solid tumors. These agents are effective against cell lines that are resistant to B-Raf and MEK1/2 inhibitor therapy. Although MK-8353 does not directly inhibit MEK1/2, it decreases the phosphorylation of ERK1/2 as well as the phosphorylation of RSK, an ERK1/2 substrate. The decrease in RSK phosphorylation appears to be a result of ERK inhibition and the decrease in ERK1/2 phosphorylation is related to the inability of MEK to catalyze the phosphorylation of the ERK-MK-8353 complex; these decreases characterize the ERK dual mechanism inhibition paradigm. Additional work will be required to determine whether ERK inhibitors will be successful in the clinic and are able to forestall the development of drug resistance of the MAP kinase pathway.

Ensemble-Based Replica Exchange Alchemical Free Energy Methods: The Effect of Protein Mutations on Inhibitor Binding

The accurate prediction of the binding affinity changes of drugs caused by protein mutations is a major goal in clinical personalized medicine. We have developed an ensemble-based free energy approach called thermodynamic integration with enhanced sampling (TIES), which yields accurate, precise, and reproducible binding affinities. TIES has been shown to perform well for predictions of free energy differences of congeneric ligands to a wide range of target proteins. We have recently introduced variants of TIES, which incorporate the enhanced sampling technique REST2 (replica exchange with solute tempering) and the free energy estimator MBAR (Bennett acceptance ratio). Here we further extend the TIES methodology to study relative binding affinities caused by protein mutations when bound to a ligand, a variant which we call TIES-PM. We apply TIES-PM to fibroblast growth factor receptor 3 (FGFR3) to investigate binding free energy changes upon protein mutations. The results show that TIES-PM with REST2 successfully captures a large conformational change and generates correct free energy differences caused by a gatekeeper mutation located in the binding pocket. Simulations without REST2 fail to overcome the energy barrier between the conformations, and hence the results are highly sensitive to the initial structures. We also discuss situations where REST2 does not improve the accuracy of predictions.

High-throughput binding affinity calculations at extreme scales

Background

Resistance to chemotherapy and molecularly targeted therapies is a major factor in limiting the effectiveness of cancer treatment. In many cases, resistance can be linked to genetic changes in target proteins, either pre-existing or evolutionarily selected during treatment. Key to overcoming this challenge is an understanding of the molecular determinants of drug binding. Using multi-stage pipelines of molecular simulations we can gain insights into the binding free energy and the residence time of a ligand, which can inform both stratified and personal treatment regimes and drug development. To support the scalable, adaptive and automated calculation of the binding free energy on high-performance computing resources, we introduce the High-throughput Binding Affinity Calculator (HTBAC). HTBAC uses a building block approach in order to attain both workflow flexibility and performance.

Results

We demonstrate close to perfect weak scaling to hundreds of concurrent multi-stage binding affinity calculation pipelines. This permits a rapid time-to-solution that is essentially invariant of the calculation protocol, size of candidate ligands and number of ensemble simulations.

Conclusions

As such, HTBAC advances the state of the art of binding affinity calculations and protocols. HTBAC provides the platform to enable scientists to study a wide range of cancer drugs and candidate ligands in order to support personalized clinical decision making based on genome sequencing and drug discovery.

Modeling the value of predictive affinity scoring in preclinical drug discovery

Drug discovery is widely recognized to be a difficult and costly activity in large part due to the challenge of identifying chemical matter which simultaneously optimizes multiple properties, one of which is affinity for the primary biological target. Further, many of these properties are difficult to predict ahead of expensive and time-consuming compound synthesis and experimental testing. Here we highlight recent work to develop compound affinity prediction models, and extensively investigate the value such models may provide to preclinical drug discovery. We demonstrate that the ability of these models to improve the overall probability of success is crucially dependent on the shape of the error distribution, not just the root-mean-square error. In particular, while scoring more molecule ideas generally improves the probability of project success when the error distribution is Gaussian, fat-tail distributions such as a Cauchy distribution, can lead to a situation where scoring more ideas actually decreases the overall probability of success.

Discovery of Potent, Selective, and Peripherally Restricted Pan-Trk Kinase Inhibitors for the Treatment of Pain

Hormones of the neurotrophin family, nerve growth factor (NGF), brain derived neurotrophic factor (BDNF), neurotrophin 3 (NT3), and neurotrophin 4 (NT4), are known to activate the family of Tropomyosin receptor kinases (TrkA, TrkB, and TrkC). Moreover, inhibition of the TrkA kinase pathway in pain has been clinically validated by the NGF antibody tanezumab, leading to significant interest in the development of small molecule inhibitors of TrkA. Furthermore, Trk inhibitors having an acceptable safety profile will require minimal brain availability. Herein, we discuss the discovery of two potent, selective, peripherally restricted, efficacious, and well-tolerated series of pan-Trk inhibitors which successfully delivered three candidate quality compounds 10b, 13b, and 19. All three compounds are predicted to possess low metabolic clearance in human that does not proceed via aldehyde oxidase-catalyzed reactions, thus addressing the potential clearance prediction liability associated with our current pan-Trk development candidate PF-06273340.

Uncertainty Quantification in Alchemical Free Energy Methods

Alchemical free energy methods have gained much importance recently from several reports of improved ligand–protein binding affinity predictions based on their implementation using molecular dynamics simulations. A large number of variants of such methods implementing different accelerated sampling techniques and free energy estimators are available, each claimed to be better than the others in its own way. However, the key features of reproducibility and quantification of associated uncertainties in such methods have barely been discussed. Here, we apply a systematic protocol for uncertainty quantification to a number of popular alchemical free energy methods, covering both absolute and relative free energy predictions. We show that a reliable measure of error estimation is provided by ensemble simulation—an ensemble of independent MD simulations—which applies irrespective of the free energy method. The need to use ensemble methods is fundamental and holds regardless of the duration of time of the molecular dynamics simulations performed.

In silico ligand-based modeling of hBACE-1 inhibitors

Alzheimer's disease is a chronic neurodegenerative disease affecting more than 30 million people worldwide. Development of small molecule inhibitors of human β-secretase 1 (hBACE-1) is being the focus of pharmaceutical industry for the past 15-20 years. Here, we successfully applied multiple ligand-based in silico modeling techniques to understand the inhibitory activities of a diverse set of small molecule hBACE-1 inhibitors reported in the scientific literature. Strikingly, the use of only a small subset of 230 (13%) molecules allowed us to develop quality models that performed reasonably well on the validation set of 1,476 (87%) inhibitors. Varying the descriptor sets and the complexity of the modeling techniques resulted in only minor improvements to the model's performance. The current results demonstrate that predictive models can be built by choosing appropriate modeling techniques in spite of using small datasets consisting of diverse chemical classes, a scenario typical in triaging of high-throughput screening results to identify false negatives. We hope that these encouraging results will help the community to develop more predictive models that would support research efforts for the debilitating Alzheimer's disease. Additionally, the integrated diversity of the techniques employed will stimulate scientists in the field to use in silico statistical modeling techniques like these to derive better models to help advance the drug discovery projects faster.

The Role of Multiscale Protein Dynamics in Antigen Presentation and T Lymphocyte Recognition

T lymphocytes are stimulated when they recognize short peptides bound to class I proteins of the major histocompatibility complex (MHC) protein, as peptide-MHC complexes. Due to the diversity in T-cell receptor (TCR) molecules together with both the peptides and MHC proteins they bind to, it has been difficult to design vaccines and treatments based on these interactions. Machine learning has made some progress in trying to predict the immunogenicity of peptide sequences in the context of specific MHC class I alleles but, as such approaches cannot integrate temporal information and lack explanatory power, their scope will always be limited. Here, we advocate a mechanistic description of antigen presentation and TCR activation which is explanatory, predictive, and quantitative, drawing on modeling approaches that collectively span several length and time scales, being capable of furnishing reliable biological descriptions that are difficult for experimentalists to provide. It is a form of multiscale systems biology. We propose the use of chemical rate equations to describe the time evolution of the foreign and host proteins to explain how the original proteins end up being presented on the cell surface as peptide fragments, while we invoke molecular dynamics to describe the key binding processes on the molecular level, including those of peptide-MHC complexes with TCRs which lie at the heart of the immune response. On each level, complementary methods based on machine learning are available, and we discuss the relationship between these divergent approaches. The pursuit of predictive mechanistic modeling approaches requires experimentalists to adapt their work so as to acquire, store, and expose data that can be used to verify and validate such models.

An Ensemble-Based Protocol for the Computational Prediction of Helix-Helix Interactions in G Protein-Coupled Receptors using Coarse-Grained Molecular Dynamics

The accurate identification of the specific points of interaction between G protein-coupled receptor (GPCR) oligomers is essential for the design of receptor ligands targeting oligomeric receptor targets. A coarse-grained molecular dynamics computer simulation approach would provide a compelling means of identifying these specific protein–protein interactions and could be applied both for known oligomers of interest and as a high-throughput screen to identify novel oligomeric targets. However, to be effective, this in silico modeling must provide accurate, precise, and reproducible information. This has been achieved recently in numerous biological systems using an ensemble-based all-atom molecular dynamics approach. In this study, we describe an equivalent methodology for ensemble-based coarse-grained simulations. We report the performance of this method when applied to four different GPCRs known to oligomerize using error analysis to determine the ensemble size and individual replica simulation time required. Our measurements of distance between residues shown to be involved in oligomerization of the fifth transmembrane domain from the adenosine A_2A receptor are in very good agreement with the existing biophysical data and provide information about the nature of the contact interface that cannot be determined experimentally. Calculations of distance between rhodopsin, CXCR4, and β₁AR transmembrane domains reported to form contact points in homodimers correlate well with the corresponding measurements obtained from experimental structural data, providing an ability to predict contact interfaces computationally. Interestingly, error analysis enables identification of noninteracting regions. Our results confirm that GPCR interactions can be reliably predicted using this novel methodology.