In silico prediction of unbound brain-to-plasma concentration ratio using machine learning algorithms

Predicting blood-brain barrier permeability of molecules with a large language model and machine learning

Predicting the blood-brain barrier (BBB) permeability of small-molecule compounds using a novel artificial intelligence platform is necessary for drug discovery. Machine learning and a large language model on artificial intelligence (AI) tools improve the accuracy and shorten the time for new drug development. The primary goal of this research is to develop artificial intelligence (AI) computing models and novel deep learning architectures capable of predicting whether molecules can permeate the human blood-brain barrier (BBB). The in silico (computational) and in vitro (experimental) results were validated by the Natural Products Research Laboratories (NPRL) at China Medical University Hospital (CMUH). The transformer-based MegaMolBART was used as the simplified molecular input line entry system (SMILES) encoder with an XGBoost classifier as an in silico method to check if a molecule could cross through the BBB. We used Morgan or Circular fingerprints to apply the Morgan algorithm to a set of atomic invariants as a baseline encoder also with an XGBoost classifier to compare the results. BBB permeability was assessed in vitro using three-dimensional (3D) human BBB spheroids (human brain microvascular endothelial cells, brain vascular pericytes, and astrocytes). Using multiple BBB databases, the results of the final in silico transformer and XGBoost model achieved an area under the receiver operating characteristic curve of 0.88 on the held-out test dataset. Temozolomide (TMZ) and 21 randomly selected BBB permeable compounds (Pred scores = 1, indicating BBB-permeable) from the NPRL penetrated human BBB spheroid cells. No evidence suggests that ferulic acid or five BBB-impermeable compounds (Pred scores < 1.29423E-05, which designate compounds that pass through the human BBB) can pass through the spheroid cells of the BBB. Our validation of in vitro experiments indicated that the in silico prediction of small-molecule permeation in the BBB model is accurate. Transformer-based models like MegaMolBART, leveraging the SMILES representations of molecules, show great promise for applications in new drug discovery. These models have the potential to accelerate the development of novel targeted treatments for disorders of the central nervous system.

Accurate prediction of Kp,uu,brain based on experimental measurement of Kp,brain and computed physicochemical properties of candidate compounds in CNS drug discovery

A mathematical equation model was developed by building the relationship between the fu,b/fu,p ratio and the computed physicochemical properties of candidate compounds, thereby predicting Kp,uu,brain based on a single experimentally measured Kp,brain value. A total of 256 compounds and 36 marketed published drugs including acidic, basic, neutral, zwitterionic, CNS-penetrant, and non-CNS penetrant compounds with diverse structures and physicochemical properties were involved in this study. A strong correlation was demonstrated between the fu,b/fu,p ratio and physicochemical parameters (CLogP and ionized fraction). The model showed good performance in both internal and external validations. The percentages of compounds with Kp,uu,brain predictions within 2-fold variability were 80.0 %-83.3 %, and more than 90 % were within a 3-fold variability. Meanwhile, "black box" QSAR models constructed by machine learning approaches for predicting fu,b/fu,p ratio based on the chemical descriptors are also presented, and the ANN model displayed the highest accuracy with an RMSE value of 0.27 and 86.7 % of the test set drugs fell within a 2-fold window of linear regression. These models demonstrated strong predictive power and could be helpful tools for evaluating the Kp,uu,brain by a single measurement parameter of Kp,brain during lead optimization for CNS penetration evaluation and ranking CNS drug candidate molecules in the early stages of CNS drug discovery.

The Trends and Future Prospective of In Silico Models from the Viewpoint of ADME Evaluation in Drug Discovery

Drug discovery and development are aimed at identifying new chemical molecular entities (NCEs) with desirable pharmacokinetic profiles for high therapeutic efficacy. The plasma concentrations of NCEs are a biomarker of their efficacy and are governed by pharmacokinetic processes such as absorption, distribution, metabolism, and excretion (ADME). Poor ADME properties of NCEs are a major cause of attrition in drug development. ADME screening is used to identify and optimize lead compounds in the drug discovery process. Computational models predicting ADME properties have been developed with evolving model-building technologies from a simplified relationship between ADME endpoints and physicochemical properties to machine learning, including support vector machines, random forests, and convolution neural networks. Recently, in the field of in silico ADME research, there has been a shift toward evaluating the in vivo parameters or plasma concentrations of NCEs instead of using predictive results to guide chemical structure design. Another research hotspot is the establishment of a computational prediction platform to strengthen academic drug discovery. Bioinformatics projects have produced a series of in silico ADME models using free software and open-access databases. In this review, we introduce prediction models for various ADME parameters and discuss the currently available academic drug discovery platforms.

QSAR model to predict Kp,uu,brain with a small dataset, incorporating predicted values of related parameter

The unbound brain-to-plasma concentration ratio (K_p,uu,brain) is a parameter that indicates the extent of central nervous system penetration. Pharmaceutical companies build prediction models because many experiments are required to obtain K_p,uu,brain. However, the lack of data hinders the design of an accurate prediction model. To construct a quantitative structure-activity relationship (QSAR) model with a small dataset of K_p,uu,brain, we investigated whether the prediction accuracy could be improved by incorporating software-predicted brain penetration-related parameters (BPrPs) as explanatory variables for pharmacokinetic parameter prediction. We collected 88 compounds with experimental K_p,uu,brain from various official publications. Random forest was used as the machine learning model. First, we developed prediction models using only structural descriptors. Second, we verified the predictive accuracy of each model with the predicted values of BPrPs incorporated in various combinations. Third, the K_p,uu,brain of the in-house compounds was predicted and compared with the experimental values. The prediction accuracy was improved using five-fold cross-validation (RMSE = 0.455, r² = 0.726) by incorporating BPrPs. Additionally, this model was verified using an external in-house dataset. The result suggested that using BPrPs as explanatory variables improve the prediction accuracy of the K_p,uu,brain QSAR model when the available number of datasets is small.

Unbound Brain-to-Plasma Partition Coefficient, Kp,uu,brain-a Game Changing Parameter for CNS Drug Discovery and Development

PURPOSE

More than 15 years have passed since the first description of the unbound brain-to-plasma partition coefficient (Kp,uu,brain) by Prof. Margareta Hammarlund-Udenaes, which was enabled by advancements in experimental methodologies including cerebral microdialysis. Since then, growing knowledge and data continue to support the notion that the unbound (free) concentration of a drug at the site of action, such as the brain, is the driving force for pharmacological responses. Towards this end, Kp,uu,brain is the key parameter to obtain unbound brain concentrations from unbound plasma concentrations.

METHODS

To understand the importance and impact of the Kp,uu,brain concept in contemporary drug discovery and development, a survey has been conducted amongst major pharmaceutical companies based in Europe and the USA. Here, we present the results from this survey which consisted of 47 questions addressing: 1) Background information of the companies, 2) Implementation, 3) Application areas, 4) Methodology, 5) Impact and 6) Future perspectives.

RESULTS AND CONCLUSIONS

From the responses, it is clear that the majority of the companies (93%) has established a common understanding across disciplines of the concept and utility of Kp,uu,brain as compared to other parameters related to brain exposure. Adoption of the Kp,uu,brain concept has been mainly driven by individual scientists advocating its application in the various companies rather than by a top-down approach. Remarkably, 79% of all responders describe the portfolio impact of Kp,uu,brain implementation in their companies as 'game-changing'. Although most companies (74%) consider the current toolbox for Kp,uu,brain assessment and its validation satisfactory for drug discovery and early development, areas of improvement and future research to better understand human brain pharmacokinetics/pharmacodynamics translation have been identified.

Prediction Models for Brain Distribution of Drugs Based on Biomimetic Chromatographic Data

The development of high-throughput approaches for the valid estimation of brain disposition is of great importance in the early drug screening of drug candidates. However, the complexity of brain tissue, which is protected by a unique vasculature formation called the blood−brain barrier (BBB), complicates the development of robust in silico models. In addition, most computational approaches focus only on brain permeability data without considering the crucial factors of plasma and tissue binding. In the present study, we combined experimental data obtained by HPLC using three biomimetic columns, i.e., immobilized artificial membranes, human serum albumin, and α1-acid glycoprotein, with molecular descriptors to model brain disposition of drugs. Kp,uu,brain, as the ratio between the unbound drug concentration in the brain interstitial fluid to the corresponding plasma concentration, brain permeability, the unbound fraction in the brain, and the brain unbound volume of distribution, was collected from literature. Given the complexity of the investigated biological processes, the extracted models displayed high statistical quality (R2 > 0.6), while in the case of the brain fraction unbound, the models showed excellent performance (R2 > 0.9). All models were thoroughly validated, and their applicability domain was estimated. Our approach highlighted the importance of phospholipid, as well as tissue and protein, binding in balance with BBB permeability in brain disposition and suggests biomimetic chromatography as a rapid and simple technique to construct models with experimental evidence for the early evaluation of CNS drug candidates.

Interpretable-ADMET: a web service for ADMET prediction and optimization based on deep neural representation

MOTIVATION

In the process of discovery and optimization of lead compounds, it is difficult for non-expert pharmacologists to intuitively determine the contribution of substructure to a particular property of a molecule.

RESULTS

In this work, we develop a user-friendly web service, named interpretable-absorption, distribution, metabolism, excretion and toxicity (ADMET), which predict 59 ADMET-associated properties using 90 qualitative classification models and 28 quantitative regression models based on graph convolutional neural network and graph attention network algorithms. In interpretable-ADMET, there are 250 729 entries associated with 59 kinds of ADMET-associated properties for 80 167 chemical compounds. In addition to making predictions, interpretable-ADMET provides interpretation models based on gradient-weighted class activation map for identifying the substructure, which is important to the particular property. Interpretable-ADMET also provides an optimize module to automatically generate a set of novel virtual candidates based on matched molecular pair rules. We believe that interpretable-ADMET could serve as a useful tool for lead optimization in drug discovery.

AVAILABILITY AND IMPLEMENTATION

Interpretable-ADMET is available at http://cadd.pharmacy.nankai.edu.cn/interpretableadmet/.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Understanding the Blood-Brain Barrier and Beyond: Challenges and Opportunities for Novel CNS Therapeutics

This review addresses questions on how to accomplish successful central nervous system (CNS) drug delivery (i.e., having the right concentration at the right CNS site, at the right time), by understanding the rate and extent of blood‐brain barrier (BBB) transport and intra‐CNS distribution in relation to CNS target site(s) exposure. To this end, we need to obtain and integrate quantitative and connected data on BBB using the Combinatory Mapping Approach that includes in vivo and ex vivo animal measurements, and the physiologically based comprehensive LEICNSPK3.0 mathematical model that can translate from animals to humans. For small molecules, slow diffusional BBB transport and active influx and efflux BBB transport determine the differences between plasma and CNS pharmacokinetics. Obviously, active efflux is important for limiting CNS drug delivery. Furthermore, liposomal formulations of small molecules may to a certain extent circumvent active influx and efflux at the BBB. Interestingly, for CNS pathologies, despite all reported disease associated BBB and CNS functional changes in animals and humans, integrative studies typically show a lack of changes on CNS drug delivery for the small molecules. In contrast, the understanding of the complex vesicle‐based BBB transport modes that are important for CNS delivery of large molecules is in progress, and their BBB transport seems to be significantly affected by CNS diseases. In conclusion, today, CNS drug delivery of small drugs can be well assessed and understood by integrative approaches, although there is still quite a long way to go to understand CNS drug delivery of large molecules.

Reproducibility of developmental neuroplasticity in in vitro brain tissue models

The current prevalence of neurodevelopmental, neurodegenerative diseases, stroke and brain injury stimulates studies aimed to identify new molecular targets, to select the drug candidates, to complete the whole set of preclinical and clinical trials, and to implement new drugs into routine neurological practice. Establishment of protocols based on microfluidics, blood-brain barrier- or neurovascular unit-on-chip, and microphysiological systems allowed improving the barrier characteristics and analyzing the regulation of local microcirculation, angiogenesis, and neurogenesis. Reconstruction of key mechanisms of brain development and even some aspects of experience-driven brain plasticity would be helpful in the establishment of brain in vitro models with the highest degree of reliability. Activity, metabolic status and expression pattern of cells within the models can be effectively assessed with the protocols of system biology, cell imaging, and functional cell analysis. The next generation of in vitro models should demonstrate high scalability, 3D or 4D complexity, possibility to be combined with other tissues or cell types within the microphysiological systems, compatibility with bio-inks or extracellular matrix-like materials, achievement of adequate vascularization, patient-specific characteristics, and opportunity to provide high-content screening. In this review, we will focus on currently available and prospective brain tissue in vitro models suitable for experimental and preclinical studies with the special focus on models enabling 4D reconstruction of brain tissue for the assessment of brain development, brain plasticity, and drug kinetics.

A curated diverse molecular database of blood-brain barrier permeability with chemical descriptors

The highly-selective blood-brain barrier (BBB) prevents neurotoxic substances in blood from crossing into the extracellular fluid of the central nervous system (CNS). As such, the BBB has a close relationship with CNS disease development and treatment, so predicting whether a substance crosses the BBB is a key task in lead discovery for CNS drugs. Machine learning (ML) is a promising strategy for predicting the BBB permeability, but existing studies have been limited by small datasets with limited chemical diversity. To mitigate this issue, we present a large benchmark dataset, B3DB, complied from 50 published resources and categorized based on experimental uncertainty. A subset of the molecules in B3DB has numerical log BB values (1058 compounds), while the whole dataset has categorical (BBB+ or BBB-) BBB permeability labels (7807). The dataset is freely available at https://github.com/theochem/B3DB and https://doi.org/10.6084/m9.figshare.15634230.v3 (version 3). We also provide some physicochemical properties of the molecules. By analyzing these properties, we can demonstrate some physiochemical similarities and differences between BBB+ and BBB- compounds.

Direct Comparison of the Prediction of the Unbound Brain-to-Plasma Partitioning Utilizing Machine Learning Approach and Mechanistic Neuropharmacokinetic Model

The mechanistic neuropharmacokinetic (neuroPK) model was established to predict unbound brain-to-plasma partitioning (K_p,uu,brain) by considering in vitro efflux activities of multiple drug resistance 1 (MDR1) and breast cancer resistance protein (BCRP). Herein, we directly compare this model to a computational machine learning approach utilizing physicochemical descriptors and efflux ratios of MDR1 and BCRP-expressing cells for predicting K_p,uu,brain in rats. Two different types of machine learning techniques, Gaussian processes (GP) and random forest regression (RF), were assessed by the time and cluster-split validation methods using 640 internal compounds. The predictivity of machine learning models based on only molecular descriptors in the time-split dataset performed worse than the cluster-split dataset, whereas the models incorporating MDR1 and BCRP efflux ratios showed similar predictivity between time and cluster-split datasets. The GP incorporating MDR1 and BCRP in the time-split dataset achieved the highest correlation (R² = 0.602). These results suggested that incorporation of MDR1 and BCRP in machine learning is beneficial for robust and accurate prediction. K_p,uu,brain prediction utilizing the neuroPK model was significantly worse compared to machine learning approaches for the same dataset. We also investigated the predictivity of K_p,uu,brain using an external independent test set of 34 marketed drugs. Compared to machine learning models, the neuroPK model showed better predictive performance with R² of 0.577. This work demonstrates that the machine learning model for K_p,uu,brain achieves maximum predictive performance within the chemical applicability domain, whereas the neuroPK model is applicable more widely beyond the chemical space covered in the training dataset.

Model-based Target Pharmacology Assessment (mTPA): An Approach Using PBPK/PD Modeling and Machine Learning to Design Medicinal Chemistry and DMPK Strategies in Early Drug Discovery

The optimal pharmacokinetic (PK) required for a drug candidate to elicit efficacy is highly dependent on the targeted pharmacology, a relationship that is often not well characterized during early phases of drug discovery. Generic assumptions around PK and potency risk misguiding screening and compound design toward nonoptimal absorption, distribution, metabolism, and excretion (ADME) or molecular properties and ultimately may increase attrition as well as hit-to-lead and lead optimization timelines. The present work introduces model-based target pharmacology assessment (mTPA), a computational approach combining physiologically based pharmacokinetic/pharmacodynamic (PBPK/PD) modeling, sensitivity analysis, and machine learning (ML) to elucidate the optimal combination of PK, potency, and ADME specific for the targeted pharmacology. Examples using frequently encountered PK/PD relationships are presented to illustrate its application, and the utility and benefits of deploying such an approach to guide early discovery efforts are discussed.

Towards Deep Neural Network Models for the Prediction of the Blood-Brain Barrier Permeability for Diverse Organic Compounds

Permeation through the blood–brain barrier (BBB) is among the most important processes controlling the pharmacokinetic properties of drugs and other bioactive compounds. Using the fragmental (substructural) descriptors representing the occurrence number of various substructures, as well as the artificial neural network approach and the double cross-validation procedure, we have developed a predictive in silico LogBB model based on an extensive and verified dataset (529 compounds), which is applicable to diverse drugs and drug-like compounds. The model has good predictivity parameters (Q2=0.815, RMSEcv=0.318) that are similar to or better than those of the most reliable models available in the literature. Larger datasets, and perhaps more sophisticated network architectures, are required to realize the full potential of deep neural networks. The analysis of fragment contributions reveals patterns of influence consistent with the known concepts of structural characteristics that affect the BBB permeability of organic compounds. The external validation of the model confirms good agreement between the predicted and experimental LogBB values for most of the compounds. The model enables the evaluation and optimization of the BBB permeability of potential neuroactive agents and other drug compounds.

The Brain Exposure Efficiency (BEE) Score

The blood-brain barrier (BBB), composed of microvascular tight junctions and glial cell sheathing, selectively controls drug permeation into the central nervous system (CNS) by either passive diffusion or active transport. Computational techniques capable of predicting molecular brain penetration are important to neurological drug design. A novel prediction algorithm, termed the Brain Exposure Efficiency Score (BEE), is presented. BEE addresses the need to incorporate the role of trans-BBB influx and efflux active transporters by considering key brain penetrance parameters, namely, steady state unbound brain to plasma ratio of drug (K_p,uu) and dose normalized unbound concentration of drug in brain (C_u,b). BEE was devised using quantitative structure-activity relationships (QSARs) and molecular modeling studies on known transporter proteins and their ligands. The developed algorithms are provided as a user-friendly open source calculator to assist in optimizing a brain penetrance strategy during the early phases of small molecule molecular therapeutic design.

ABC Transporters at the Blood-Brain Interfaces, Their Study Models, and Drug Delivery Implications in Gliomas

Drug delivery into the brain is regulated by the blood-brain interfaces. The blood-brain barrier (BBB), the blood-cerebrospinal fluid barrier (BCSFB), and the blood-arachnoid barrier (BAB) regulate the exchange of substances between the blood and brain parenchyma. These selective barriers present a high impermeability to most substances, with the selective transport of nutrients and transporters preventing the entry and accumulation of possibly toxic molecules, comprising many therapeutic drugs. Transporters of the ATP-binding cassette (ABC) superfamily have an important role in drug delivery, because they extrude a broad molecular diversity of xenobiotics, including several anticancer drugs, preventing their entry into the brain. Gliomas are the most common primary tumors diagnosed in adults, which are often characterized by a poor prognosis, notably in the case of high-grade gliomas. Therapeutic treatments frequently fail due to the difficulty of delivering drugs through the brain barriers, adding to diverse mechanisms developed by the cancer, including the overexpression or expression de novo of ABC transporters in tumoral cells and/or in the endothelial cells forming the blood-brain tumor barrier (BBTB). Many models have been developed to study the phenotype, molecular characteristics, and function of the blood-brain interfaces as well as to evaluate drug permeability into the brain. These include in vitro, in vivo, and in silico models, which together can help us to better understand their implication in drug resistance and to develop new therapeutics or delivery strategies to improve the treatment of pathologies of the central nervous system (CNS). In this review, we present the principal characteristics of the blood-brain interfaces; then, we focus on the ABC transporters present on them and their implication in drug delivery; next, we present some of the most important models used for the study of drug transport; finally, we summarize the implication of ABC transporters in glioma and the BBTB in drug resistance and the strategies to improve the delivery of CNS anticancer drugs.

Investigation of MDR1-overexpressing cell lines to derive a quantitative prediction approach for brain disposition using in vitro efflux activities

MDR1-overexpressing Lilly Laboratories cell porcine kidney 1 cells (LLC-PK1-MDR1) and Madin-Darby canine kidney cells (MDCK-MDR1) are widely used in drug discovery to evaluate the in vivo relevance of MDR1-mediated efflux. However, as the in vitro efflux ratio (ER) of these cell lines are variable among research facilities, the in vitro ER of these cell lines that would affect quantitative predictivity of brain disposition has not been fully clarified. The aim of this study was to examine the effect of ER on the quantitative predictivity of brain disposition toward compounds with MDR1 and/or breast cancer resistant protein (BCRP) liabilities. Test compounds including internal molecules and five typical substrates of MDR1 and/or BCRP were assessed via an in vitro transporter assay to determine the corrected flux ratio (CFR) and an in vivo animal study using wild-type (WT) and Mdr1a (-/-)/Bcrp(-/-) (dual KO) rats. To assess the in vivo ER for MDR1, the two cell lines LLC-PK1-MDR1 and MDCK-MDR1 were used. After intravenously administering 29 test compounds to rats, the K_p,brain ratio (ratio of K_p,brain,WT to K_{p,brain,dual KO}), which is considered to be the unbound plasma-to-brain ratio (K_p,uu,brain) that does not require correction for protein binding in both plasma and brain, was determined by measuring their concentrations in the plasma and brain. The K_p,brain ratio of these compounds was predicted by fitting scaling factor that was extrapolated from the in vitro to in vivo ER for MDR1 and BCRP, defined as α and β, respectively. K_p,brain ratio values of 83% and 68% of compounds were predicted by using MDCK-MDR1 and LLC-PK1-MDR1, respectively, within a 2-fold range of the actual corresponding values. The α predicted from CFRs of MDCK-MDR1 was 47-fold smaller than that of LLC-PK1-MDR1; however, a dramatic change in β was not observed. This result appears to be consistent with the data of in vitro transport activity of MDR1, which was estimated to be ~28-fold higher in MDCK-MDR1 than in LLC-PK1-MDR1 by correlation analysis with CFR. Through this study, we revealed that 1) brain disposition in rats was well-predicted by considering the in vitro efflux activities for both MDR1 and BCRP, and 2) MDCK-MDR1 was the superior cell line for the quantitative prediction of brain disposition.

Prediction of brain:blood unbound concentration ratios in CNS drug discovery employing in silico and in vitro model systems

Recent years have seen a paradigm shift away from optimizing the brain:blood concentration ratio toward the more relevant brain:blood unbound concentration ratio (K_p,uu,br) in CNS drug discovery. Here, we review the recent developments in the in silico and in vitro model systems to predict the K_p,uu,br of discovery compounds with special emphasis on the in-vitro-in-vivo correlation. We also discuss clinical 'translation' of rodent K_p,uu,br and highlight the future directions for improvement in brain penetration prediction. Important in this regard are in silico K_p,uu,br models built on larger datasets of high quality, calibration and deeper understanding of experimental in vitro transporter systems, and better understanding of blood-brain barrier transporters and their in vivo relevance aside from P-gp and BCRP.

Predicting Drug Concentration-Time Profiles in Multiple CNS Compartments Using a Comprehensive Physiologically-Based Pharmacokinetic Model

Drug development targeting the central nervous system (CNS) is challenging due to poor predictability of drug concentrations in various CNS compartments. We developed a generic physiologically based pharmacokinetic (PBPK) model for prediction of drug concentrations in physiologically relevant CNS compartments. System‐specific and drug‐specific model parameters were derived from literature and in silico predictions. The model was validated using detailed concentration‐time profiles from 10 drugs in rat plasma, brain extracellular fluid, 2 cerebrospinal fluid sites, and total brain tissue. These drugs, all small molecules, were selected to cover a wide range of physicochemical properties. The concentration‐time profiles for these drugs were adequately predicted across the CNS compartments (symmetric mean absolute percentage error for the model prediction was <91%). In conclusion, the developed PBPK model can be used to predict temporal concentration profiles of drugs in multiple relevant CNS compartments, which we consider valuable information for efficient CNS drug development.

Optimization of physicochemical properties for 4-anilinoquinazoline inhibitors of trypanosome proliferation

Human African trypanosomiasis (HAT) is a deadly disease in need of new chemotherapeutics that can cross into the central nervous system. We previously reported the discovery of 2 (NEU-617), a small molecule with activity against T. brucei bloodstream proliferation. Further optimization of 2 to improve the physicochemical properties (LogP, LLE, [1], and MPO score) [2] have led us to twelve sub-micromolar compounds, most importantly the headgroup variants 9i and 9j, and the linker variant 18. Although these 3 compounds had reduced potency compared to 2, they all had improved LogP, LLE and MPO scores. Cross-screening these analogs against other protozoan parasites uncovered 9o with potent activity towards T. brucei, T. cruzi and L. major, while four others compounds (17, 18, 21, 26) showed activity towards P. falciparum D6. This reinforces the effectiveness of lead repurposing for the discovery of new protozoan disease therapeutics.

Novel CNS drug discovery and development approach: model-based integration to predict neuro-pharmacokinetics and pharmacodynamics

INTRODUCTION

CNS drug development has been hampered by inadequate consideration of CNS pharmacokinetic (PK), pharmacodynamics (PD) and disease complexity (reductionist approach). Improvement is required via integrative model-based approaches. Areas covered: The authors summarize factors that have played a role in the high attrition rate of CNS compounds. Recent advances in CNS research and drug discovery are presented, especially with regard to assessment of relevant neuro-PK parameters. Suggestions for further improvements are also discussed. Expert opinion: Understanding time- and condition dependent interrelationships between neuro-PK and neuro-PD processes is key to predictions in different conditions. As a first screen, it is suggested to use in silico/in vitro derived molecular properties of candidate compounds and predict concentration-time profiles of compounds in multiple compartments of the human CNS, using time-course based physiology-based (PB) PK models. Then, for selected compounds, one can include in vitro drug-target binding kinetics to predict target occupancy (TO)-time profiles in humans. This will improve neuro-PD prediction. Furthermore, a pharmaco-omics approach is suggested, providing multilevel and paralleled data on systems processes from individuals in a systems-wide manner. Thus, clinical trials will be better informed, using fewer animals, while also, needing fewer individuals and samples per individual for proof of concept in humans.

A Generic Multi-Compartmental CNS Distribution Model Structure for 9 Drugs Allows Prediction of Human Brain Target Site Concentrations

Purpose

Predicting target site drug concentration in the brain is of key importance for the successful development of drugs acting on the central nervous system. We propose a generic mathematical model to describe the pharmacokinetics in brain compartments, and apply this model to predict human brain disposition.

Methods

A mathematical model consisting of several physiological brain compartments in the rat was developed using rich concentration-time profiles from nine structurally diverse drugs in plasma, brain extracellular fluid, and two cerebrospinal fluid compartments. The effect of active drug transporters was also accounted for. Subsequently, the model was translated to predict human concentration-time profiles for acetaminophen and morphine, by scaling or replacing system- and drug-specific parameters in the model.

Results

A common model structure was identified that adequately described the rat pharmacokinetic profiles for each of the nine drugs across brain compartments, with good precision of structural model parameters (relative standard error <37.5%). The model predicted the human concentration-time profiles in different brain compartments well (symmetric mean absolute percentage error <90%).

Conclusions

A multi-compartmental brain pharmacokinetic model was developed and its structure could adequately describe data across nine different drugs. The model could be successfully translated to predict human brain concentrations.

Electronic supplementary material

The online version of this article (doi:10.1007/s11095-016-2065-3) contains supplementary material, which is available to authorized users.

QSAR Model of Unbound Brain-to-Plasma Partition Coefficient, Kp,uu,brain: Incorporating P-glycoprotein Efflux as a Variable

We report development and prospective validation of a QSAR model of the unbound brain-to-plasma partition coefficient, K_p,uu,brain, based on the in-house data set of ∼1000 compounds. We discuss effects of experimental variability, explore the applicability of both regression and classification approaches, and evaluate a novel, model-within-a-model approach of including P-glycoprotein efflux prediction as an additional variable. When tested on an independent test set of 91 internal compounds, incorporation of P-glycoprotein efflux information significantly improves the model performance resulting in an R² of 0.53, RMSE of 0.57, Spearman's Rho correlation coefficient of 0.73, and qualitative prediction accuracy of 0.8 (kappa = 0.6). In addition to improving the performance, one of the key advantages of this approach is the larger chemical space coverage provided indirectly through incorporation of the in vitro, higher throughput data set that is 4 times larger than the in vivo data set.

Recent progresses in the exploration of machine learning methods as in-silico ADME prediction tools

In-silico methods have been explored as potential tools for assessing ADME and ADME regulatory properties particularly in early drug discovery stages. Machine learning methods, with their ability in classifying diverse structures and complex mechanisms, are well suited for predicting ADME and ADME regulatory properties. Recent efforts have been directed at the broadening of application scopes and the improvement of predictive performance with particular focuses on the coverage of ADME properties, and exploration of more diversified training data, appropriate molecular features, and consensus modeling. Moreover, several online machine learning ADME prediction servers have emerged. Here we review these progresses and discuss the performances, application prospects and challenges of exploring machine learning methods as useful tools in predicting ADME and ADME regulatory properties.

Towards Better BBB Passage Prediction Using an Extensive and Curated Data Set

In the present report, the challenging task of drug delivery across the blood-brain barrier (BBB) is addressed via a computational approach. The BBB passage was modeled using classification and regression schemes on a novel extensive and curated data set (the largest to the best of our knowledge) in terms of log BB. Prior to the model development, steps of data analysis that comprise chemical data curation, structural, cutoff and cluster analysis (CA) were conducted. Linear Discriminant Analysis (LDA) and Multiple Linear Regression (MLR) were used to fit classification and correlation functions. The best LDA-based model showed overall accuracies over 85 % and 83 % for the training and test sets, respectively. Also a MLR-based model with acceptable explanation of more than 69 % of the variance in the experimental log BB was developed. A brief and general interpretation of proposed models allowed the estimation on how 'near' our computational approach is to the factors that determine the passage of molecules through the BBB. In a final effort some popular and powerful Machine Learning methods were considered. Comparable or similar performance was observed respect to the simpler linear techniques. Most of the compounds with anomalous behavior were put aside into a set denoted as controversial set and discussion regarding to these compounds is provided. Finally, our results were compared with methodologies previously reported in the literature showing comparable to better results. The results could represent useful tools available and reproducible by all scientific community in the early stages of neuropharmaceutical drug discovery/development projects.

Virtual screening, ADMET profiling, molecular docking and dynamics approaches to search for potent selective natural molecules based inhibitors against metallothionein-III to study Alzheimer's disease

MOTIVATION

Metallothionein-III (MT-III) displays neuro-inhibitory activity and is involved in the repair of neuronal damage. An altered expression level of MT-III suggests that it could be a mitigating factor in Alzheimer's disease (AD) neuronal dysfunction. Currently there are limited marketed drugs available against MT-III. The inhibitors are mostly pseudo-peptide based with limited ADMET. In our present study, available database InterBioScreen (natural compounds) was screened out for MT-III. Pharmacodynamics and pharmacokinetic studies were performed. Molecular docking and simulations of top hit molecules were performed to study complex stability.

RESULTS

Study reveals potent selective molecules that interact and form hydrogen bonds with amino acids Ser-6 and Lys-22 are common to established melatonin inhibitors for MT-III. These include DMHMIO, MCA B and s27533 derivatives. The ADMET profiling was better with comparable interaction energy values. It includes properties like blood brain barrier, hepatotoxicity, druggability, mutagenicity and carcinogenicity. Molecular dynamics studies were performed to validate our findings.

Molecular properties determining unbound intracellular and extracellular brain exposure of CNS drug candidates

In the present work we sought to gain a mechanistic understanding of the physicochemical properties that influence the transport of unbound drug across the blood-brain barrier (BBB) as well as the intra- and extracellular drug exposure in the brain. Interpretable molecular descriptors that significantly contribute to the three key neuropharmacokinetic properties related to BBB drug transport (Kp,uu,brain), intracellular accumulation (Kp,uu,cell), and binding and distribution in the brain (Vu,brain) for a set of 40 compounds were identified using partial least-squares (PLS) analysis. The tailoring of drug properties for improved brain exposure includes decreasing the polarity and/or hydrogen bonding capacity. The design of CNS drug candidates with intracellular targets may benefit from an increase in basicity and/or the number of hydrogen bond donors. Applying this knowledge in drug discovery chemistry programs will allow designing compounds with more desirable CNS pharmacokinetic properties.

In silico prediction of brain exposure: drug free fraction, unbound brain to plasma concentration ratio and equilibrium half-life

The focus of CNS drug pharmacokinetics programs has recently shifted from determining the total concentrations in brain and blood to considering also unbound fractions and concentrations. Unfortunately, assessing unbound brain exposure experimentally requires demanding in vivo and in vitro studies. We propose a physical model, based on lipid binding and pH partitioning, to predict in silico the unbound volume of distribution in the brain. The model takes into account the partition of a drug into lipids, interstitial fluid and intracellular compartments of the brain. The results are in good agreement with the experimental data, suggesting that the contributions of lipid binding and pH partitioning are important in determining drug exposure in brain. The predicted values are used, together with predictions for plasma protein binding, as corrective terms in a second model to derive the unbound brain to plasma concentration ratio starting from experimental values of total concentration ratio. The calculated values of brain free fraction and passive permeability are also used to qualitatively determine the brain to plasma equilibration time in a model that shows promising results but is limited to a very small set of compounds. The models we propose are a step forward in understanding and predicting pharmacologically relevant exposure in brain starting from compounds' chemical structure and neuropharmacokinetics, by using experimental total brain to plasma ratios, in silico calculated properties and simple physics-based approaches. The models can be used in central nervous system drug discovery programs for a fast and cheap assessment of unbound brain exposure. For existing compounds, the unbound ratios can be derived from experimental values of total brain to plasma ratios. For both existing and hypothetical compounds, the unbound volume of distribution due to lipid binding and pH partitioning can be calculated starting only from the chemical structure.

Computer modeling of blood brain barrier permeability of physiologically active compounds

At present work discusses the current level of computer modeling the relationship structure of organic compounds and drugs and their ability to penetrate the BBB. All descriptors that influence to this permeability within classification and regression QSAR models are generalized and analyzed. The crucial role of H-bond in processes both passive, and active transport across BBB is observed. It is concluded that further research should be focused on interpretation the spatial structure of a full-size P-glycoprotein molecule with high resolution and the creation of QSAR models describing the quantitative relationship between structure and active transport of substances across BBB.

In silico methods to predict drug toxicity

This review describes in silico methods to characterize the toxicity of pharmaceuticals, including tools which predict toxicity endpoints such as genotoxicity or organ-specific models, tools addressing ADME processes, and methods focusing on protein-ligand docking binding. These in silico tools are rapidly evolving. Nowadays, the interest has shifted from classical studies to support toxicity screening of candidates, toward the use of in silico methods to support the expert. These methods, previously considered useful only to provide a rough, initial estimation, currently have attracted interest as they can assist the expert in investigating toxic potential. They provide the expert with safety perspectives and insights within a weight-of-evidence strategy. This represents a shift of the general philosophy of in silico methodology, and it is likely to further evolve especially exploiting links with system biology.

Improving the prediction of drug disposition in the brain

INTRODUCTION

Ability to cross the blood-brain barrier is one of the key ADME characteristics of all drug candidates regardless of their target location in the body. While good brain penetration is essential for CNS drugs, it may lead to serious side effects in case of peripherally-targeted molecules. Despite a high demand of computational methods for estimating brain transport early in drug discovery, achieving good prediction accuracy still remains a challenging task.

AREAS COVERED

This article reviews various measures employed to quantify brain delivery and recent advances in QSAR approaches for predicting these properties from the compound's structure. Additionally, the authors discuss the classification models attempting to distinguish between permeable and impermeable chemicals.

EXPERT OPINION

Recent research in the field of brain penetration modeling showed an increasing understanding of the processes involved in drug disposition, although most models of brain/plasma partitioning still rely on purely statistical considerations. Preferably, new models should incorporate mechanistic knowledge since it is the prerequisite for guiding drug design efforts in the desired direction. To increase the efficiency of computational tools, a broader view is necessary, involving rate and extent of brain penetration, as well as plasma and brain tissue binding strength, instead of relying on any single property.

Computational prediction of CNS drug exposure based on a novel in vivo dataset

PURPOSE

To develop a computational model for predicting CNS drug exposure using a novel in vivo dataset.

METHODS

The brain-to-plasma (B:P) ratio of 43 diverse compounds was assessed following intravenous administration to Swiss Outbred mice. B:P ratios were subjected to PLS modeling using calculated molecular descriptors. The obtained results were transferred to a qualitative setting in which compounds predicted to have a B:P ratio > 0.3 were sorted as high CNS exposure compounds and those below this value were sorted as low CNS exposure compounds. The model was challenged with an external test set consisting of 251 compounds for which semi-quantitative values of CNS exposure were available in the literature.

RESULTS

The dataset ranged more than 1700-fold in B:P ratio, with 16 and 27 compounds being sorted as low and high CNS exposure drugs, respectively. The model was a one principal component model based on five descriptors reflecting molecular shape, electronegativity, polarisability and charge transfer, and allowed 74% of the compounds in the training set and 76% of the test set to be predicted correctly.

CONCLUSION

A qualitative computational model has been developed which accurately classifies compounds as being high or low CNS exposure drugs based on rapidly calculated molecular descriptors.

Getting the MAX out of Computational Models: The Prediction of Unbound-Brain and Unbound-Plasma Maximum Concentrations

The objective of this work was to establish that unbound maximum concentrations may be reasonably predicted from a combination of computed molecular properties assuming subcutaneous (SQ) dosing. Additionally, we show that the maximum unbound plasma and brain concentrations may be projected from a mixture of in vitro absorption, distribution, metabolism, excretion experimental parameters in combination with computed properties (volume of distribution, fraction unbound in microsomes). Finally, we demonstrate the utility of the underlying equations by showing that the maximum total plasma concentrations can be projected from the experimental parameters for a set of compounds with data collected from clinical research.