In silico approaches and tools for the prediction of drug metabolism and fate: A review

https://doi.org/10.1016/j.compbiomed.2019.01.008Get rights and content

Highlights

  • In silico approaches and tools for predicting drug metabolism and fate are reviewed.

  • QSAR, machine learning, and computational docking/molecular dynamics are described.

  • Computational models for predicting metabolic enzymatic reactions are summarized.

  • A list of tools for in silico prediction is provided.

  • Concerns and limitations of predictive model construction are outlined.

Abstract

The fate of administered drugs is largely influenced by their metabolism. For example, endogenous enzyme–catalyzed conversion of drugs may result in therapeutic inactivation or activation or may transform the drugs into toxic chemical compounds. This highlights the importance of drug metabolism in drug discovery and development, and accounts for the wide variety of experimental technologies that provide insights into the fate of drugs. In view of the high cost of traditional drug development, a number of computational approaches have been developed for predicting the metabolic fate of drug candidates, allowing for screening of large numbers of chemical compounds and then identifying a small number of promising candidates. In this review, we introduce in silico approaches and tools that have been developed to predict drug metabolism and fate, and assess their potential to facilitate the virtual discovery of promising drug candidates. We also provide a brief description of various recent models for predicting different aspects of enzyme-drug reactions and provide a list of recent in silico tools used for drug metabolism prediction.

Introduction

As our understanding of drug fate–determining metabolic reactions has increased, drug metabolism has attracted greater attention as a critical factor in drug discovery [[1], [2], [3]]. The fate of substances such as drugs and xenobiotics administered into our body is largely governed by the three phases of drug metabolism: phase I, introduction of a reactive group by oxidation, reduction or hydrolysis, among others; phase II, conjugation with various moieties; and phase III, removal of xenobiotics and metabolites from cells in the liver and intestine. These transformation processes may convert compounds to inactive, active, or toxic metabolites. Not surprisingly, because it is responsible for the clearance of ∼70% of clinical drugs, metabolism has been intensively investigated as part of drug development efforts [4].

Natural compounds have recently attracted considerable research attention owing to their inherent advantages and high potential as drug candidates [5]. Moreover, the structural similarity of certain natural compounds with metabolites found in the human body makes metabolism a critical factor in determining the effectiveness of natural drugs [[6], [7], [8]]. For instance, historical opioid drug candidates are metabolized into more potent metabolites, such as (dihydro) codeine, which, in turn, is metabolized into (dihydro) morphine [9,10]. Considering the large number of endogenous enzymatic reactions that influence drug modification through (de)activation and (de)toxification, determining how a drug is metabolized is an important step in drug discovery.

Numerous experimental technologies have been used in recent decades to study the metabolism and fate of drugs [[11], [12], [13]]. The traditional drug discovery method—target-to-hit, hit-to-lead, and lead optimization—is expensive, costing more than $200 million for the average drug, and time-consuming, with a typical discovery period of 4–5 years [14,15]. In addition, because it is impossible to exactly replicate biological in vivo environments, such methods are relatively inaccurate and are still considered low-throughput, given the scale of combinatorial structural variations of chemical compounds.

Numerous advances in predicting drug metabolism using in silico approaches have been made as part of drug discovery efforts, and different aspects of these advances have been reviewed [[16], [17], [18], [19], [20], [21], [22]]. These include tools to predict drug metabolism based on the interactions of drugs with cytochrome P450 (CYP450) enzymes and their metabolic endpoints [18,22], tools to predict ADMET (absorption, distribution, metabolism, excretion, and toxicity) properties of drugs and related solubility, permeability and bioavailability issues [16], as well as approaches for predicting the inducibility of drug-metabolizing enzymes and transporters that affect the plasma concentration of drugs, which can cause undesirable or prolonged action or adverse effects [21].

Given these observations, in silico approaches have been increasingly used to predict the metabolic conversion of drugs [23] and as such are considered the best “fail early and fail cheap strategy,” allowing for reduced costs, time savings, and thus decreased attrition rates in late drug discovery phases. Herein, we present an introduction to fundamental approaches to in silico model development for predicting the metabolic fate of drugs as well as their toxicity. In addition, we summarize currently available in silico tools and some recent achievements in predicting drug metabolism. Finally, we discuss the future and challenges of in silico drug discovery.

Section snippets

Predictions based on quantitative structure-activity relationships and machine-learning approaches

The quantitative structure-activity relationship (QSAR) concept, developed in the early 1960s by Hansch/Fujita [24] and Free/Wilson [25] and widely used in drug discovery, assumes that molecules with similar structures potentially exhibit similar chemical and biological activities [[26], [27], [28]]. The original concept of structure-activity relationship dates back to 1868, when Crum-Brown and Fraser introduced the idea of correlating the chemical composition of a certain compound with its

Prediction of drug conversion to toxic metabolites

The metabolism of xenobiotics such as drugs and other foreign substances involves certain important enzymatic reactions, such as those mediated by CYP450 family enzymes expressed in liver and small intestines. According to literature reports, ∼90% of drugs can be efficiently metabolized by six CYP450 enzymes [104], the activity of which can be altered by factors such as genetic polymorphisms, cytokine regulation, disease state, sex, age, and hormones [[105], [106], [107]]. Another example is

Problems associated with predictive model construction

The inconsistency of available experimental data used to build in silico models is a major concern [119]. Predictive models strongly rely on experimental data for model construction; thus, high variability in experimental assays produced by biological variation and technical errors can lead to erroneous data and may therefore introduce inaccuracy into predictive models. The inaccuracy of in silico models may also result from different experimental conditions for the multiple resources

Conclusions

Because of its central importance, metabolism in biological systems has been intensively researched, especially in the field of drug discovery. The high impact of drug metabolism on drug efficacy and drug fate in biological systems has given rise to numerous in silico approaches and tools for metabolic reaction prediction in recent decades. However, the limitations of these approaches cannot be ignored. Specifically, the fact that these methods are highly dependent on experimental data is a

Conflicts of interest

The authors declare no conflicts of interest.

Acknowledgements and grants

This research was supported by a National Research Foundation of Korea grant funded by the Korean government (NRF-2016R1D1A1B03935264) and was also supported by the Bio-Synergy Research Project (NRF-2018M3A9C4076474) of the Ministry of Science, ICT and Future Planning through the National Research Foundation.

References (124)

  • Y. Cong et al.

    Quantitative structure–activity relationship study of influenza virus neuraminidase A/PR/8/34 (H1N1) inhibitors by genetic algorithm feature selection and support vector regression

    Chemom. Intell. Lab. Syst.

    (2013)
  • L. Zhang et al.

    From machine learning to deep learning: progress in machine intelligence for rational drug discovery

    Drug Discov. Today

    (2017)
  • Y.-C. Lo et al.

    Machine learning in chemoinformatics and drug discovery

    Drug Discov. Today

    (2018)
  • Y. Sakiyama et al.

    Predicting human liver microsomal stability with machine learning techniques

    J. Mol. Graph. Model.

    (2008)
  • H. Li et al.

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

    J. Pharm. Sci.

    (2007)
  • H. Chen et al.

    The rise of deep learning in drug discovery, Drug Discov

    Today Off.

    (2018)
  • C.W. Yap et al.

    Quantitative structure–pharmacokinetic relationships for drug clearance by using statistical learning methods

    J. Mol. Graph. Model.

    (2006)
  • A. Roncaglioni et al.

    In silico methods to predict drug toxicity

    Curr. Opin. Pharmacol.

    (2013)
  • A. Lavecchia

    Machine-learning approaches in drug discovery: methods and applications, Drug Discov

    Today Off.

    (2015)
  • A. Ganesan et al.

    Molecular dynamics-driven drug discovery: leaping forward with confidence

    Drug Discov. Today

    (2017)
  • G. Jones et al.

    Development and validation of a genetic algorithm for flexible docking

    J. Mol. Bio.

    (1997)
  • C. B-Rao et al.

    Managing protein flexibility in docking and its applications, Drug Discov

    Today Off.

    (2009)
  • H.A. Carlson

    Protein flexibility and drug design: how to hit a moving target

    Curr. Opin. Chem. Biol.

    (2002)
  • Y.Z. Chen et al.

    Prediction of potential toxicity and side effect protein targets of a small molecule by a ligand-protein inverse docking approach

    J. Mol. Graph. Model.

    (2001)
  • Z. Yao et al.

    Using molecular docking-based binding energy to predict toxicity of binary mixture with different binding sites

    Chemosphere

    (2013)
  • Z.L. Ji et al.

    In silico search of putative adverse drug reaction related proteins as a potential tool for facilitating drug adverse effect prediction

    Toxicol. Lett.

    (2006)
  • M.A. Lazar et al.

    Physiology. De-meaning of metabolism

    Science

    (2012)
  • R.A. Cairns et al.

    Regulation of cancer cell metabolism

    Nat. Rev. Cancer

    (2011)
  • Y.P. Kang et al.

    Recent advances in cancer metabolism: a technological perspective

    Exp. Mol. Med.

    (2018)
  • G.B. Stefano et al.

    Endogenous morphine: up-to-date review 2011

    Folia Biol

    (2012)
  • L.C. Kirkwood et al.

    Characterization of the human cytochrome P450 enzymes involved in the metabolism of dihydrocodeine

    Br. J. Clin. Pharmacol.

    (1997)
  • X. Diao et al.

    In vitro and in vivo human metabolism of a new synthetic cannabinoid NM-2201 (CBL-2201)

    Forensic Toxicol.

    (2017)
  • S.A. Wrighton et al.

    In vitro methods for assessing human hepatic drug metabolism: their use in drug development

    Drug Metab. Rev.

    (1993)
  • S.M. Paul et al.

    How to improve R&D productivity: the pharmaceutical industry's grand challenge

    Nat. Rev. Drug Discovery

    (2010)
  • S. Alqahtani

    In silico ADME-Tox modeling: progress and prospects

    Expert Opin. Drug Metab. Toxicol.

    (2017)
  • C.H. Andrade et al.

    In silico prediction of drug metabolism by P450

    Curr. Drug Metab.

    (2014)
  • B.H. Tan et al.

    In vitro and in silico approaches to study cytochrome P450-mediated interactions

    J. Pharm. Sci.

    (2017)
  • T. Zhang et al.

    In silico prediction of cytochrome P450-mediated drug metabolism

    Comb. Chem. High Throughput Screening

    (2011)
  • M.P. Gleeson et al.

    In-silico ADME models: a general assessment of their utility in drug discovery applications

    Curr. Top. Med. Chem.

    (2011)
  • D. Schuster et al.

    Predicting drug metabolism induction in silico

    Curr. Top. Med. Chem.

    (2006)
  • T. Fox et al.

    Machine learning techniques for in silico modeling of drug metabolism

    Curr. Top. Med. Chem.

    (2006)
  • J.K. Lu Tan

    Software for metabolism prediction

  • C. Hansch et al.

    p-σ-π analysis. A method for the correlation of biological activity and chemical structure

    J. Am. Chem. Soc.

    (1964)
  • S.M. Free et al.

    A mathematical contribution to structure-activity studies

    J. Med. Chem.

    (1964)
  • J.C. Gertrudes et al.

    Machine learning techniques and drug design

    Curr. Med. Chem.

    (2012)
  • Q.S. Du et al.

    Recent advances in QSAR and their applications in predicting the activities of chemical molecules, peptides and proteins for drug design

    Curr. Protein Pept. Sci.

    (2008)
  • S.H. Oraphan Phuangsawai et al.

    Quantitative structure–activity relationship (QSAR) methods for the prediction of substrates, inhibitors, and inducers of metabolic enzymes

  • A.C. Brown et al.

    On the connection between chemical constitution and physiological action; with special reference to the physiological action of the salts of the ammonium bases derived from strychnia, brucia, thebaia, codeia, morphia, and nicotia

    J. Anat. Physiol.

    (1868)
  • S. Kausar et al.

    An automated framework for QSAR model building

    J. Cheminform.

    (2018)
  • G. Domenico et al.

    Applicability domain for QSAR models: where theory meets reality

    International journal of quantitative structure-property relationships (IJQSPR)

    (2016)
  • Cited by (76)

    • Exploration of anticancer potential of Lantadenes from weed Lantana camara: Synthesis, in silico, in vitro and in vivo studies

      2023, Phytochemistry
      Citation Excerpt :

      The prognostic outcomes from Table S3 specified that except for compounds 1 & 5 al l the molecules indicated the non-targeting potential towards the brain and signify the safe profile of Lantadene scaffold. Further, the fate of administered drug is also influenced by its metabolism, that involves enzymatic modification of the molecule into their active/inactive metabolites depicting their therapeutic response (Kazmi et al., 2019). Possible metabolism of Lantadenes (1–2), reduced Lantadenes (3–4) and their derivatives (5–10) with P450 CYP3A4 substrate, P450 CYP3A4 inhibitor and P450 CYP2C9 inhibitor have been expressed with positive and negative symbols.

    View all citing articles on Scopus
    1

    These authors contributed equally.

    View full text