Machine learning approaches for predicting compounds that interact with therapeutic and ADMET related proteins

Computational methods for predicting compounds of specific pharmacodynamic and ADMET (absorption, distribution, metabolism, excretion and toxicity) property are useful for facilitating drug discovery and evaluation. Recently, machine learning methods such as neural networks and support vector machines have been explored for predicting inhibitors, antagonists, blockers, agonists, activators and substrates of proteins related to specific therapeutic and ADMET property. These methods are particularly useful for compounds of diverse structures to complement QSAR methods, and for cases of unavailable receptor 3D structure to complement structure-based methods. A number of studies have demonstrated the potential of these methods for predicting such compounds as substrates of P-glycoprotein and cytochrome P450 CYP isoenzymes, inhibitors of protein kinases and CYP isoenzymes, and agonists of serotonin receptor and estrogen receptor. This article is intended to review the strategies, current progresses and underlying difficulties in using machine learning methods for predicting these protein binders and as potential virtual screening tools. Algorithms for proper representation of the structural and physicochemical properties of compounds are also evaluated.     

Bayesian neural nets for modeling in drug discovery

Bayesian regularized artificial neural networks (BRANNs) are used in the development of quantitative SAR models. These networks have the potential to solve several problems that arise in QSAR modeling such as choice of model, robustness of model, choice of validation set, size of validation effort, and optimization of network architecture. The application of the methods to a wide range of problems, including target-based QSAR, ADMET modeling and eukaryotic promoter finding, is illustrated.     

Cellular models for ADMET predictions and evaluation of drug-drug interactions

Deficiencies in ADMET (absorption, distribution, metabolism, excretion and toxicity) properties and drug-drug interactions are collectively the major causes of attrition during drug development. As such, assays have been developed with which to study and optimize these key properties in early dug discovery. While screening using systems expressing discrete proteins have provided valuable insight, quantitative structure-activity relationships (QSARs) and predictive computational models, the ability to study several processes in tandem is paramount to in vivo projection. In particular, the key role of transporter proteins in controlling access to drug metabolizing enzymes and other intracellular processes cannot be overlooked. In this respect, cellular models provide a key platform to study the complex interplay between xenobiotic transport and metabolism, which underlie many ADMET issues. In addition, uptake and accumulation in tissues may provide a mechanistic insight into false negatives arising from simple, primary screens, for example, cytochrome P450 (CYP) inhibition analysis. Qualitative and quantitative interspecies differences in the regulation, expression and functional activity of key ADMET processes confound extrapolation from animals to man. However, complementary screens using animal and human material may assist the interpretation of safety assessment findings and help project the risk for early human studies.     

Synthesis, antimicrobial evaluation, QSAR and in Silico ADMET studies of decanoic acid derivatives

Various derivatives of decanoic acid (CD) have been synthesized and evaluated against Gram positive B. subtilis, S. aureus and Gram negative E. coli bacteria as well a sagainst fungi C. albicans and A. niger. Quantitative structure activity relationship (QSAR) models for antimicrobial activities were developed using multiple linear regression and cross validated by leave one out (LOO) approach. QSAR studies indicated that activity against Gram positive bacteria was governed by lipophilicity of the compounds while topologicalsteric nature of the molecule was deciding factor for antifungal activity. Further, in silico ADMET studies showed that compounds CD12, 19, 20 and 23 could be explored further for other activities.     

Exploiting QSAR models in lead optimization

QSAR models can play a vital role in both the opening phase and the endgame of lead optimization. In the opening phase, there is often a large quantity of data from high-throughput screening (HTS), and potential leads need to be selected from several distinct chemotypes. In the endgame, the throughput of the final, critical ADMET and pharmacokinetic assays is often not sufficient to allow full experimental characterization of all the structures in the available time. A considerable amount of the current research toward new QSAR models is based on the modeling of the general ADMET phenomena, with the aim of constructing globally applicable models. The process to construct QSAR models is relatively straightforward; however, it is also simple to build misleading, or even incorrect, models. This review considers the key developments in the field of QSAR modeling: how QSAR models are constructed, how they can be validated, their reliability and their applicability. If applied carefully and appropriately, the QSAR technique has a valuable role to play during lead optimization.     

Combined structure-based pharmacophore, virtual screening, and 3D-QSAR studies of structural diverse dehydrosqualene synthase inhibitors

Dehydrosqualene synthase (CrtM) is a key enzyme in the synthesis of presqualene diphosphate in Staphylococcus aureus. In the current study, a combination of structure-based pharmacophore and 3D-QSAR methods are used to clarify the essential quantitative structure–activity relationship (QSAR) of known CrtM inhibitors; the multicomplex-based pharmacophore (MCBP) guided method has been suggested to generate a comprehensive pharmacophore of CrtM based on twenty crystal structures of CrtM inhibitor complex. Performances of the MCBP-based virtual screening approach were applied to screen specs chemical databases (202, 408 compounds). Thirty-eight compounds were selected from the final hits and should be shifted to experimental studies. The MCBP model has been successfully used to identify the bioactive conformation and align 24 structurally diverse CrtM inhibitors. The QSAR analyses have been performed on these CrtM inhibitors based on MCBP guided alignment. These results may provide important information for further design and discovery of novel CrtM inhibitors.     

In silico approaches for predicting ADME properties of drugs

Combinatorial chemistry and high-throughput screening have increased the possibility of finding new lead compounds at much shorter time periods than conventional medicinal chemistry. However, too much promising drug candidates often fail because of unsatisfactory ADME properties. In silico ADME studies are expected to reduce the risk of late-stage attrition of drug development and to optimize screening and testing by looking at only the promising compounds. To this end, many in silico approaches for predicting ADME properties of compounds from their chemical structure have been developed, ranging from data-based approaches such as quantitative structure-activity relationship (QSAR), similarity searches, and 3-dimensional QSAR, to structure-based methods such as ligand-protein docking and pharmacophore modelling. In addition, several methods of integrating ADME properties to predict pharmacokinetics at the organ or body level have been studied. In this article, we briefly summarize in silico ADME approaches.     

Ligand-based structural hypotheses for virtual screening

The majority of drug targets for small molecule therapeutics are proteins whose three-dimensional structure is not known to sufficient resolution to permit structure-based design. All three-dimensional QSAR approaches have a requirement for some hypothesis of ligand conformation and alignment, and predictions of molecular activity critically depend on this ligand-based binding site hypothesis. The molecular similarity function used in the Surflex docking system, coupled with quantitative pressure to minimize overall molecular volume, forms an effective objective function for generating hypotheses of bioactive conformations of sets of small molecules binding to their cognate proteins. Results are presented, assessing utility of the method for ligands of the serotonin, histamine, muscarinic, and GABA(A) receptors. The Surflex similarity module (Surflex-Sim) was able, in each case, to distinguish true ligands from random compounds using models constructed from just two or three known ligands. True positive rates of 60% were achieved with false positive rates of 0-3%; the theoretical enrichment rates were over 150-fold compared with random screening. The methods are practically applicable for rational design of ligands and for high-throughput virtual screening and offer competitive performance to many structure-based docking algorithms.     

3D QSAR and docking studies of a series of histone deacetylase inhibitors

In recent years, a class of intracellular metalloproteases known as histone deacetylases (HDACs) has become popular targets for cancer therapy. HDACs play an important role in the modification of chromatin structure and regulation of gene expression. In this study, structure-based and ligand-based methods are used to provide a theoretical basis for finding highly potent antitumor drugs. 61 small molecules were studied by three-dimensional quantitative structure–activity relationship (3D QSAR) method. Comparative molecular field analysis and comparative molecular similarity indices analysis methods were employed to build the 3D QSAR model of HDAC inhibitors containing hydroxamic acid group. As the 3D-structure of target HDACs has been investigated by the X-ray crystallographic studies, the binding mode between compounds and HDACs can be explored by docking approach.     

Modulation of P-glycoprotein (PGP) mediated multidrug resistance (MDR) using chemosensitizers: recent advances in the design of selective MDR modulators

Over the past two decades, a number of chemical entities have been investigated in the continuing quest to reverse P-glycoprotein (PGP) mediated multidrug resistance (MDR) in cancer. The complexity of interactions between these agents and the proteins responsible for MDR in conjunction with the challenges associated with developing SAR/QSAR relationships for MDR modulators has hampered our ability to develop agents that modulate MDR with enhanced specificity of target, increased efficacy, and minimized toxicity when coadministered with anticancer drugs. With an increased understanding of the molecular interaction, target-mediated SAR and combinatorial chemistry approaches, newer more selective inhibitors have been recently reported. These agents have shown remarkable promise in preclinical trials; although their ultimate clinical therapeutic utility remains to be established. The emphasis of this review is placed on the current understanding of modulator-drug transport protein interactions and to review the advances in the structure-based design, synthetic efforts and the cellular pharmacology of MDR modulating activity of a number of known PGP inhibitors.     

Utility of protein structures in overcoming ADMET-related issues of drug-like compounds

The number of solved X-ray structures of proteins relevant for ADMET processes of drug molecules has increased remarkably over recent years. In principle, this development offers the possibility to complement the quantitative structure-property relationship (QSPR)-dominated repertoire of in silico ADMET methods with protein-structure-based approaches. However, the complex nature and the weak nonspecific ligand-binding properties of ADMET proteins take structural biology methods and current docking programs to the limit. In this review we discuss the utility of protein-structure-based design and docking approaches aimed at overcoming issues related to plasma protein binding, active transport via P-glycoprotein, hERG channel mediated cardiotoxicity and cytochrome P450 inhibition, metabolism and induction.     

Prediction of drug–ABC-transporter interaction — Recent advances and future challenges

With the discovery of P-glycoprotein (P-gp), it became evident that ABC-transporters play a vital role in bioavailability and toxicity of drugs. They prevent intracellular accumulation of toxic compounds, which renders them a major defense mechanism against xenotoxic compounds. Their expression in cells of all major barriers (intestine, blood–brain barrier, blood–placenta barrier) as well as in metabolic organs (liver, kidney) also explains their influence on the ADMET properties of drugs and drug candidates. Thus, in silico models for the prediction of the probability of a compound to interact with P-gp or analogous transporters are of high value in the early phase of the drug discovery process. Within this review, we highlight recent developments in the area, with a special focus on the molecular basis of drug–transporter interaction. In addition, with the recent availability of X-ray structures of several ABC-transporters, also structure-based design methods have been applied and will be addressed.     

Improving structure-based virtual screening by multivariate analysis of scoring data

Three different multivariate statistical methods, PLS discriminant analysis, rule-based methods, and Bayesian classification, have been applied to multidimensional scoring data from four different target proteins: estrogen receptor alpha (ERalpha), matrix metalloprotease 3 (MMP3), factor Xa (fXa), and acetylcholine esterase (AChE). The purpose was to build classifiers able to discriminate between active and inactive compounds, given a structure-based virtual screen. Seven different scoring functions were used to generate the scoring matrices. The classifiers were compared to classical consensus scoring and single scoring functions. The classifiers show a superior performance, with rule-based methods being most effective. The precision of correctly predicting an active compound is about 90% for three of the targets and about 25% for acetylcholine esterase. On the basis of these results, a new two-stage approach is suggested for structure-based virtual screening where limited activity information is available.     

药物渗透性预测方法的研究进展

药物渗透性(透膜能力)对药物生物利用度影响重大,在药物研发过程中先导化合物的渗透性影响其成药几率。很多药物由于渗透性不佳而遭撤市,因此在药物研发早期对先导化合物进行渗透性预测,可以加速药物研发进程、节约研发经费。通过对现有预测药物渗透性的理论方法(定量构效关系、自由能差法、分子动力学模拟等)和有关软件(ADMET Predictor、GastroPlus、QikProp、Volsurf等)进行介绍,并评价了其优缺点,以期根据需要选择合适理论方法,提高预测的准确性。