Systems-ADME/Tox: resources and network approaches

Machine Learning Models Identify New Inhibitors for Human OATP1B1

The uptake transporter OATP1B1 (SLC01B1) is largely localized to the sinusoidal membrane of hepatocytes and is a known victim of unwanted drug-drug interactions. Computational models are useful for identifying potential substrates and/or inhibitors of clinically relevant transporters. Our goal was to generate OATP1B1 in vitro inhibition data for [3H] estrone-3-sulfate (E3S) transport in CHO cells and use it to build machine learning models to facilitate a comparison of seven different classification models (Deep learning, Adaboosted decision trees, Bernoulli naïve bayes, k-nearest neighbors (knn), random forest, support vector classifier (SVC), logistic regression (lreg), and XGBoost (xgb)] using ECFP6 fingerprints to perform 5-fold, nested cross validation. In addition, we compared models using 3D pharmacophores, simple chemical descriptors alone or plus ECFP6, as well as ECFP4 and ECFP8 fingerprints. Several machine learning algorithms (SVC, lreg, xgb, and knn) had excellent nested cross validation statistics, particularly for accuracy, AUC, and specificity. An external test set containing 207 unique compounds not in the training set demonstrated that at every threshold SVC outperformed the other algorithms based on a rank normalized score. A prospective validation test set was chosen using prediction scores from the SVC models with ECFP fingerprints and were tested in vitro with 15 of 19 compounds (84% accuracy) predicted as active (≥20% inhibition) showed inhibition. Of these compounds, six (abamectin, asiaticoside, berbamine, doramectin, mobocertinib, and umbralisib) appear to be novel inhibitors of OATP1B1 not previously reported. These validated machine learning models can now be used to make predictions for drug-drug interactions for human OATP1B1 alongside other machine learning models for important drug transporters in our MegaTrans software.

Structural Relationship Study of Octanol-Water Partition Coefficient of Some Sulfa Drugs Using GA-MLR and GA-ANN Methods

AIM AND OBJECTIVE

Sulfonamides (sulfa drugs) are compounds with a wide range of biological activities and they are the basis of several groups of drugs. Quantitative Structure-Property Relationship (QSPR) models are derived to predict the logarithm of water/ 1-octanol partition coefficients (logP) of sulfa drugs.

MATERIALS AND METHODS

A data set of 43 sulfa drugs was randomly divided into 3 groups: training, test and validation sets consisting of 70%, 15% and 15% of data point, respectively. A large number of molecular descriptors were calculated with Dragon software. The Genetic Algorithm - Multiple Linear Regressions (GA-MLR) and genetic algorithm -artificial neural network (GAANN) were employed to design the QSPR models. The possible molecular geometries of sulfa drugs were optimized at B3LYP/6-31G* level with Gaussian 98 software. The molecular descriptors derived from the Dragon software were used to build a predictive model for prediction logP of mentioned compounds. The Genetic Algorithm (GA) method was applied to select the most relevant molecular descriptors.

RESULTS

The R2 and MSE values of the MLR model were calculated to be 0.312 and 5.074 respectively. R2 coefficients were 0.9869, 0.9944 and 0.9601for the training, test and validation sets of the ANN model, respectively.

CONCLUSION

Comparison of the results revealed that the application the GA-ANN method gave better results than GA-MLR method.

Comparison of Deep Learning With Multiple Machine Learning Methods and Metrics Using Diverse Drug Discovery Data Sets

Machine learning methods have been applied to many data sets in pharmaceutical research for several decades. The relative ease and availability of fingerprint type molecular descriptors paired with Bayesian methods resulted in the widespread use of this approach for a diverse array of end points relevant to drug discovery. Deep learning is the latest machine learning algorithm attracting attention for many of pharmaceutical applications from docking to virtual screening. Deep learning is based on an artificial neural network with multiple hidden layers and has found considerable traction for many artificial intelligence applications. We have previously suggested the need for a comparison of different machine learning methods with deep learning across an array of varying data sets that is applicable to pharmaceutical research. End points relevant to pharmaceutical research include absorption, distribution, metabolism, excretion, and toxicity (ADME/Tox) properties, as well as activity against pathogens and drug discovery data sets. In this study, we have used data sets for solubility, probe-likeness, hERG, KCNQ1, bubonic plague, Chagas, tuberculosis, and malaria to compare different machine learning methods using FCFP6 fingerprints. These data sets represent whole cell screens, individual proteins, physicochemical properties as well as a data set with a complex end point. Our aim was to assess whether deep learning offered any improvement in testing when assessed using an array of metrics including AUC, F1 score, Cohen's kappa, Matthews correlation coefficient and others. Based on ranked normalized scores for the metrics or data sets Deep Neural Networks (DNN) ranked higher than SVM, which in turn was ranked higher than all the other machine learning methods. Visualizing these properties for training and test sets using radar type plots indicates when models are inferior or perhaps over trained. These results also suggest the need for assessing deep learning further using multiple metrics with much larger scale comparisons, prospective testing as well as assessment of different fingerprints and DNN architectures beyond those used.

In silico methods for predicting drug-drug interactions with cytochrome P-450s, transporters and beyond

Drug-drug interactions (DDIs) are associated with severe adverse effects that may lead to the patient requiring alternative therapeutics and could ultimately lead to drug withdrawal from the market if they are severe. To prevent the occurrence of DDI in the clinic, experimental systems to evaluate drug interaction have been integrated into the various stages of the drug discovery and development process. A large body of knowledge about DDI has also accumulated through these studies and pharmacovigillence systems. Much of this work to date has focused on the drug metabolizing enzymes such as cytochrome P-450s as well as drug transporters, ion channels and occasionally other proteins. This combined knowledge provides a foundation for a hypothesis-driven in silico approach, using either cheminformatics or physiologically based pharmacokinetics (PK) modeling methods to assess DDI potential. Here we review recent advances in these approaches with emphasis on hypothesis-driven mechanistic models for important protein targets involved in PK-based DDI. Recent efforts with other informatics approaches to detect DDI are highlighted. Besides DDI, we also briefly introduce drug interactions with other substances, such as Traditional Chinese Medicines to illustrate how in silico modeling can be useful in this domain. We also summarize valuable data sources and web-based tools that are available for DDI prediction. We finally explore the challenges we see faced by in silico approaches for predicting DDI and propose future directions to make these computational models more reliable, accurate, and publically accessible.

Open Source Bayesian Models. 1. Application to ADME/Tox and Drug Discovery Datasets

On the order of hundreds of absorption, distribution, metabolism, excretion, and toxicity (ADME/Tox) models have been described in the literature in the past decade which are more often than not inaccessible to anyone but their authors. Public accessibility is also an issue with computational models for bioactivity, and the ability to share such models still remains a major challenge limiting drug discovery. We describe the creation of a reference implementation of a Bayesian model-building software module, which we have released as an open source component that is now included in the Chemistry Development Kit (CDK) project, as well as implemented in the CDD Vault and in several mobile apps. We use this implementation to build an array of Bayesian models for ADME/Tox, in vitro and in vivo bioactivity, and other physicochemical properties. We show that these models possess cross-validation receiver operator curve values comparable to those generated previously in prior publications using alternative tools. We have now described how the implementation of Bayesian models with FCFP6 descriptors generated in the CDD Vault enables the rapid production of robust machine learning models from public data or the user’s own datasets. The current study sets the stage for generating models in proprietary software (such as CDD) and exporting these models in a format that could be run in open source software using CDK components. This work also demonstrates that we can enable biocomputation across distributed private or public datasets to enhance drug discovery.

Fusion of nonclinical and clinical data to predict human drug safety

Adverse drug reactions continue to be a major cause of morbidity in both patients receiving therapeutics and in drug R&D programs. Predicting and possibly eliminating these adverse events remains a high priority in industry, government agencies and healthcare systems. With small molecule candidates, the fusion of nonclinical and clinical data is essential in establishing an overall system that creates a true translational science approach. Several new advances are taking place that attempt to create a 'patient context' mechanism early in drug research and development and ultimately into the marketplace. This 'life-cycle' approach has as its core the development of human-oriented, nonclinical end points and the incorporation of clinical knowledge at the drug design stage. The next 5 years should witness an explosion of what the author views as druggable and safe chemical space, pharmacosafety molecular targets and the most important aspect, an understanding of unique susceptibilities in patients developing adverse drug reactions. Our current knowledge of clinical safety relies completely on pharmacovigilance data from approved and marketed drugs, with a few exceptions of drugs failing in clinical trials. Massive data repositories now and soon to be available via cloud computing should stimulate a major effort in expanding our view of clinical drug safety and its incorporation into early drug research and development.

Biomarkers

Biomarkers are characteristics objectively measured and evaluated as indicators of: normal biologic processes, pathogenic processes, or pharmacologic response(s) to a therapeutic intervention. In environmental research and risk assessment, biomarkers are frequently referred to as indicators of human or environmental hazards. Discovering and implementing new biomarkers for toxicity caused by exposure to a chemical either from a therapeutic intervention or accidentally through the environment continues to be pursued through the use of animal models to predict potential human effects, from human studies (clinical or epidemiologic) or biobanked human samples, or the combination of all such approaches. The key to discovering or inferring biomarkers through computational means involves the identification or prediction of the molecular target(s) of the chemical(s) and the association of these targets with perturbed biological pathways. Two examples are given in this chapter: (1) inferring potential human biomarkers from animal toxicogenomics data, and (2) the identification of protein targets through computational means and associating these in one example with potential drug interactions and in another case with increasing the risk of developing certain human diseases.

Molecular docking and virtual screening for novel protein tyrosine phosphatase 1B (PTP1B) inhibitors

Protein tyrosine phosphatase 1B (PTP1B) functions as major negative regulator of insulin and leptin signaling pathways. In view of this, PTP1B is an significant target for drug development against cancer, diabetes and obesity. The aim of the current study is to identify PTP1B inhibitors by means of virtual screening with docking. 523,366 molecules from ZINC database have been screened and based on DOCK grid scores and hydrogen bonding interactions five new potential inhibitors were identified. ZINC12502589, ZINC13213457, ZINC25721858, ZINC31392733 and ZINC04096400 were identified as potential lead molecules for inhibition of PTP1B. The identified molecules were subjected to Lipinski's rule of five parameters and found that they did not violate any rule. More specific analysis of pharmacological parameters may be scrutinized through a complete ADME/Tox evaluation. Pharma algorithm was used to Calculate ADME-Tox profiles for such molecules. In general, all the molecules presented advantages and as well as disadvantages when compared to each other. No marked difference in health effects and toxicity profiles were observed among these molecules.

Virtual screening and evaluation of Ketol-Acid Reducto-Isomerase (KARI) as a putative drug target for Aspergillosis

Aspergillus is a leading causative agent for fungal morbidity and mortality in immuno-compromised patients. To identify a putative target to design or identify new antifungal drug, against Aspergillus is required. In our previous work, we have analyzed the various biochemical pathways, and we found Ketol Acid Reducto-Isomerase (KARI) an enzyme involves in the amino acid biosynthesis, could be a better target. This enzyme was found to be unique by comparing to host proteome through BLASTp analysis. A homology based model of KARI was generated by Swiss model server. The generated model had been validated by PROCHECK and WHAT IF programs. The Zinc library was generated within the limitation of the Lipinski rule of five, for docking study. Based on the dock-score six molecules have been studied for ADME/TOX analysis and subjected for pharmacophore model generation. The Zinc ID of the potential inhibitors is ZINC00720614, ZINC01068126, ZINC0923, ZINC02090678, ZINC00663057 and ZINC02284065 and found to be pharmacologically active agonist and antagonist of KARI. This study is an attempt to Insilco evaluation of the KARI as a drug target and the screened inhibitors could help in the development of the better drug against Aspergillus.

Accessing, using, and creating chemical property databases for computational toxicology modeling

Toxicity data is expensive to generate, is increasingly seen as precompetitive, and is frequently used for the generation of computational models in a discipline known as computational toxicology. Repositories of chemical property data are valuable for supporting computational toxicologists by providing access to data regarding potential toxicity issues with compounds as well as for the purpose of building structure-toxicity relationships and associated prediction models. These relationships use mathematical, statistical, and modeling computational approaches and can be used to understand the mechanisms by which chemicals cause harm and, ultimately, enable prediction of adverse effects of these chemicals to human health and/or the environment. Such approaches are of value as they offer an opportunity to prioritize chemicals for testing. An increasing amount of data used by computational toxicologists is being published into the public domain and, in parallel, there is a greater availability of Open Source software for the generation of computational models. This chapter provides an overview of the types of data and software available and how these may be used to produce predictive toxicology models for the community.

Integration of pharmacogenetics and pharmacogenomics in drug development: implications for regulatory and medical decision making in pediatric diseases

This article aims to provide an overview of the current situation regarding pharmacogenetic and pharmacogenomic (PG) studies in pediatrics, with a special focus on the role of PG data in the regulatory decision-making process. Despite the gap in pharmacogenetic research due to the lack of translational studies in adults and children, several technologies exist in drug development and biomarkers validation, which could supply valuable information concerning labeling and dosing recommendations. If performed under strict good clinical practice quality criteria, such findings could be included in the submission package of new chemical entities and used as additional information for prescribers, supporting further evaluation and understanding of the efficacy and safety profile of new medicines. Even though regulatory authorities may be aware of the potential role of PG in medical practice and guidances are available about the integration of PG in drug development, most data obtained from PG studies are not used by prescribers. The challenge is to better understand whether PG markers can be used to assess potential differences in drug response during the clinical program, so PG data can be integrated into the regulatory decision-making process, enabling the introduction of labeling information that promotes optimal dosing in the pediatric population.

Evolution of promiscuous nuclear hormone receptors: LXR, FXR, VDR, PXR, and CAR

Nuclear hormone receptors (NHRs) are transcription factors that work in concert with co-activators and co-repressors to regulate gene expression. Some examples of ligands for NHRs include endogenous compounds such as bile acids, retinoids, steroid hormones, thyroid hormone, and vitamin D. This review describes the evolution of liver X receptors α and β (NR1H3 and 1H2, respectively), farnesoid X receptor (NR1H4), vitamin D receptor (NR1I1), pregnane X receptor (NR1I2), and constitutive androstane receptor (NR1I3). These NHRs participate in complex, overlapping transcriptional regulation networks involving cholesterol homeostasis and energy metabolism. Some of these receptors, particularly PXR and CAR, are promiscuous with respect to the structurally wide range of ligands that act as agonists. A combination of functional and computational analyses has shed light on the evolutionary changes of NR1H and NR1I receptors across vertebrates, and how these receptors may have diverged from ancestral receptors that first appeared in invertebrates.

In silico repositioning of approved drugs for rare and neglected diseases

One approach to speed up drug discovery is to examine new uses for existing approved drugs, so-called 'drug repositioning' or 'drug repurposing', which has become increasingly popular in recent years. Analysis of the literature reveals many examples of US Food and Drug Administration-approved drugs that are active against multiple targets (also termed promiscuity) that can also be used to therapeutic advantage for repositioning for other neglected and rare diseases. Using proof-of-principle examples, we suggest here that with current in silico technologies and databases of the structures and biological activities of chemical compounds (drugs) and related data, as well as close integration with in vitro screening data, improved opportunities for drug repurposing will emerge for neglected or rare/orphan diseases.

Toxicokinetic Considerations in Predicting Toxicity

The ability of a compound to elicit a toxic effect within an organism is dependent upon three factors (i) the external exposure of the organism to the toxicant in the environment or via the food chain (ii) the internal uptake of the compound into the organism and its transport to the site of action in sufficient concentration and (iii) the inherent toxicity of the compound. The in silico prediction of toxicity and the role of external exposure have been dealt with in other chapters of this book. This chapter focuses on the importance of ‘internal exposure’ i.e. the absorption, distribution, metabolism and elimination (ADME) properties of compounds which determine their toxicokinetic profile. An introduction to key concepts in toxicokinetics will be provided, along with examples of modelling approaches and software available to predict these properties. A brief introduction will also be given into the theory of physiologically-based toxicokinetic modelling.

Toxicity testing in the 21st century: a vision and a strategy

With the release of the landmark report Toxicity Testing in the 21st Century: A Vision and a Strategy, the U.S. National Academy of Sciences, in 2007, precipitated a major change in the way toxicity testing is conducted. It envisions increased efficiency in toxicity testing and decreased animal usage by transitioning from current expensive and lengthy in vivo testing with qualitative endpoints to in vitro toxicity pathway assays on human cells or cell lines using robotic high-throughput screening with mechanistic quantitative parameters. Risk assessment in the exposed human population would focus on avoiding significant perturbations in these toxicity pathways. Computational systems biology models would be implemented to determine the dose-response models of perturbations of pathway function. Extrapolation of in vitro results to in vivo human blood and tissue concentrations would be based on pharmacokinetic models for the given exposure condition. This practice would enhance human relevance of test results, and would cover several test agents, compared to traditional toxicological testing strategies. As all the tools that are necessary to implement the vision are currently available or in an advanced stage of development, the key prerequisites to achieving this paradigm shift are a commitment to change in the scientific community, which could be facilitated by a broad discussion of the vision, and obtaining necessary resources to enhance current knowledge of pathway perturbations and pathway assays in humans and to implement computational systems biology models. Implementation of these strategies would result in a new toxicity testing paradigm firmly based on human biology.

Drug effects viewed from a signal transduction network perspective

Understanding how drugs affect cellular network structures and how resulting signals are translated into drug effects holds the key to the discovery of medicines. Herein we examine this cause-effect relationship by determining protein network structures associated with the generation of specific in vivo drug-effect patterns. Medicines having similar in vivo pharmacology have been identified by a comparison of drug-effect profiles of 1320 medicines. Protein network positions reached by these medicines were ascertained by examining the coinvestigation frequency of these medicines and 1179 protein network constituents in millions of scientific investigations. Interestingly, medicine associations obtained by comparing by drug-effect profiles mirror those obtained by comparing drug-protein coinvestigation frequency profiles, demonstrating that these drug-protein reachability profiles are relevant to in vivo pharmacology. By using protein associations obtained in these investigations and independent, curated protein interaction information, drug-mediated protein network topology models can be constructed. These protein network topology models reveal that drugs having similar pharmacology profiles reach similar discrete positions in cellular protein network systems and provide a network view of medicine cause-effect relationships.

Can systems toxicology identify common biomarkers of non-genotoxic carcinogenesis?

For the rapid development of safe, efficacious chemicals it is important that any potential liabilities are identified as early as possible in the discovery/development pipeline. Once identified it is then possible to make rational decisions on whether to progress a chemical and/or series further; one such liability is chemical carcinogenesis, a highly undesirable characteristic in a novel chemical entity. Chemical carcinogens may be roughly divided into two classes, those that elicit their actions through direct damage to DNA (genotoxic carcinogens) and those that cause carcinogenesis through mechanisms that involve direct damage of the DNA by the agent (non-genotoxic carcinogens). Whereas the former group can be identified by in vitro screens to a good degree of accuracy, the latter group are far more problematic due to their diverse modes of action. This review will focus on the latter class of chemical carcinogens, examining how modern '-omic' technologies have begun to identify signatures that may represent sensitive, early markers for these processes. In addition to their use in signature generation the role of -omic level approaches to delineating molecular mechanisms of action will also be discussed.

Carrier-mediated cellular uptake of pharmaceutical drugs: an exception or the rule?

It is generally thought that many drug molecules are transported across biological membranes via passive diffusion at a rate related to their lipophilicity. However, the types of biophysical forces involved in the interaction of drugs with lipid membranes are no different from those involved in their interaction with proteins, and so arguments based on lipophilicity could also be applied to drug uptake by membrane transporters or carriers. In this article, we discuss the evidence supporting the idea that rather than being an exception, carrier-mediated and active uptake of drugs may be more common than is usually assumed - including a summary of specific cases in which drugs are known to be taken up into cells via defined carriers - and consider the implications for drug discovery and development.

Future directions for drug transporter modelling

Since the late 1980s computational methods such as quantitative structure-activity relationship (QSAR) and pharmacophore approaches have become more widely applied to assess interactions between drug-like molecules and transporters, starting with P-glycoprotein (P-gp). Identifying molecules that interact with P-gp and other transporters is important for drug discovery, but it is normally ascertained using laborious in vitro and in vivo studies. Computational QSAR and pharmacophore models can be used to screen commercial databases of molecules rapidly and suggest those likely to bind as substrates or inhibitors for transporters. These predictions can then be readily verified in vitro, thus representing a more efficient route to screening. Recently, the application of this approach has seen the identification of new substrates and inhibitors for several transporters. The successful application of computational models and pharmacophore models in particular to predict transporter binding accurately represents a way to anticipate drug-drug interactions of novel molecules from molecular structure. These models may also see incorporation in future pharmacokinetic-pharmacodynamic models to improve predictions of in vivo drug effects in patients. The implications of early assessment of transporter activity, current advances in QSAR, and other computational methods for future development of these and systems-based approaches will be discussed.

The renaissance of H/D exchange

The increasing demand for stable isotopically labeled compounds has led to an increased interest in H/D-exchange reactions at carbon centers. Today deuterium-labeled compounds are used as internal standards in mass spectrometry or to help elucidate mechanistic theories. Access to these deuterated compounds takes place significantly more efficiently and more cost effectively by exchange of hydrogen by deuterium in the target molecule than by classical synthesis. This Review will concentrate on the preparative application of the H/D-exchange reaction in the preparation of deuterium-labeled compounds. Advances over the last ten years are brought together and critically evaluated.

In silico pharmacology for drug discovery: methods for virtual ligand screening and profiling

Pharmacology over the past 100 years has had a rich tradition of scientists with the ability to form qualitative or semi-quantitative relations between molecular structure and activity in cerebro. To test these hypotheses they have consistently used traditional pharmacology tools such as in vivo and in vitro models. Increasingly over the last decade however we have seen that computational (in silico) methods have been developed and applied to pharmacology hypothesis development and testing. These in silico methods include databases, quantitative structure-activity relationships, pharmacophores, homology models and other molecular modeling approaches, machine learning, data mining, network analysis tools and data analysis tools that use a computer. In silico methods are primarily used alongside the generation of in vitro data both to create the model and to test it. Such models have seen frequent use in the discovery and optimization of novel molecules with affinity to a target, the clarification of absorption, distribution, metabolism, excretion and toxicity properties as well as physicochemical characterization. The aim of this review is to illustrate some of the in silico methods for pharmacology that are used in drug discovery. Further applications of these methods to specific targets and their limitations will be discussed in the second accompanying part of this review.

Algorithms for network analysis in systems-ADME/Tox using the MetaCore and MetaDrug platforms

The authors have previously applied two integrated platforms, MetaCore and MetaDrug, for the assembly and analysis of human biological networks as a useful method for the integration and functional interpretation of high-throughput experimental data. The present study demonstrates in detail the specific algorithms that are used in both software platforms. Using a standard set of genes as input, namely CYP3A4 (an enzyme), PXR (a nuclear hormone receptor), MDR1 (a transporter) and hERG (an ion channel) related to the absorption, distribution, metabolism, excretion and toxicity (ADME/Tox) of xenobiotics, we have now generated networks with each algorithm. The relative advantages and disadvantages of these algorithms are explained using these examples as well as appropriate instances of utility to illustrate further the particular circumstances for their use. In addition, the benefits of the different network algorithms are identified when compared with algorithms available in other products, where this information is available.

Methods for Predicting Human Drug Metabolism

Drug metabolism information is a necessary component of drug discovery and development. The key issues in drug metabolism include identifying: the enzyme(s) involved, the site(s) of metabolism, the resulting metabolite(s), and the rate of metabolism. Methods for predicting human drug metabolism from in vitro and computational methodologies and determining relationships between the structure and metabolic activity of molecules are also critically important for understanding potential drug interactions and toxicity. There are numerous experimental and computational approaches that have been developed in order to predict human metabolism which have their own limitations. It is apparent that few of the computational tools for metabolism prediction alone provide the major integrated functions needed to assist in drug discovery. Similarly the different in vitro methods for human drug metabolism themselves have implicit limitations. The utilization of these methods for pharmaceutical and other applications as well as their integration is discussed as it is likely that hybrid methods will provide the most success.

Computer-aided drug discovery and development (CADDD): in silico-chemico-biological approach

It is generally recognized that drug discovery and development are very time and resources consuming processes. There is an ever growing effort to apply computational power to the combined chemical and biological space in order to streamline drug discovery, design, development and optimization. In biomedical arena, computer-aided or in silico design is being utilized to expedite and facilitate hit identification, hit-to-lead selection, optimize the absorption, distribution, metabolism, excretion and toxicity profile and avoid safety issues. Commonly used computational approaches include ligand-based drug design (pharmacophore, a 3D spatial arrangement of chemical features essential for biological activity), structure-based drug design (drug-target docking), and quantitative structure-activity and quantitative structure-property relationships. Regulatory agencies as well as pharmaceutical industry are actively involved in development of computational tools that will improve effectiveness and efficiency of drug discovery and development process, decrease use of animals, and increase predictability. It is expected that the power of CADDD will grow as the technology continues to evolve.

Application of data mining approaches to drug delivery

Computational approaches play a key role in all areas of the pharmaceutical industry from data mining, experimental and clinical data capture to pharmacoeconomics and adverse events monitoring. They will likely continue to be indispensable assets along with a growing library of software applications. This is primarily due to the increasingly massive amount of biology, chemistry and clinical data, which is now entering the public domain mainly as a result of NIH and commercially funded projects. We are therefore in need of new methods for mining this mountain of data in order to enable new hypothesis generation. The computational approaches include, but are not limited to, database compilation, quantitative structure activity relationships (QSAR), pharmacophores, network visualization models, decision trees, machine learning algorithms and multidimensional data visualization software that could be used to improve drug delivery after mining public and/or proprietary data. We will discuss some areas of unmet needs in the area of data mining for drug delivery that can be addressed with new software tools or databases of relevance to future pharmaceutical projects.