Current Topics in Medicinal Chemistry - Volume 8, Issue 18, 2008

Nowadays, it is well known that Quantitative studies on Structure-Activity and Structure-Property relationships (QSAR/QSPR) may become powerful tools to help Medicinal Chemistry scientist in directed drug research. In past years, various strategies have been developed to characterize and classify structural patterns of low weighted drugs by means of molecular descriptors. It has become possible not only to assess diversities or similarities of structure databases, but molecular descriptors also facilitate the identification of potential bioactive molecules from the rapidly increasing number of compound libraries. They even allow for a controlled de-novo design of new lead structures. The number and apparent diversity of molecular descriptors develop in this sense is wide covering from constitutional, and physicochemical properties to 3D descriptors. For instance, the Handbook of Molecular Descriptors (HMD) published by Todeschini and Consoni describe more than 1 500 molecular descriptors grouped on more than 15 different families. This is the most comprehensive collection of molecular descriptors and presents a detailed review from the origins of this research field up to present day. This practically oriented reference book gives a thorough overview of the different molecular descriptors representations and their corresponding molecular descriptors. Anyhow, the research in this field is far from ended and more recently, in the post-HMD times, are being investigated new trends in the development of molecular descriptors and computational techniques to seek QSAR models as well. Specifically, the class of Topological Indices (TIs), previously collected in HMD, has called the attention of many researchers due to their important capacity to capture biologically relevant information but being very simple and fast to calculate. New descriptors appeared that come to reinforce the tool of indices published in HMD but also go beyond these frontiers opening new possibilities such as the study of proteins, RNAs, Complex Networks, which is a more broad view for TIs that have been developed in parallel to small-sized drug indices. Specifically, the application of TIs or modified variants of them was used in the past to study social and technological Complex Networks and are being now “rediscovered” for molecular sciences. That is way, the present collection of papers pretend to humbly call the attention of experimental and theoretical authors (of different and somehow parallel areas) on the new scenarios that may be considered when we put in the same bag all together Medicinal Chemistry, Drug Design, Proteomics, QSAR, Complex Systems theory, and Bioinformatics. At the same time, we pretend to give some introductory but useful details on the particularities of legal issues that appear in Medicinal Chemistry research when we involve QSAR and Bioinformatics such as: copyright protection, patents registry, trademarks, and taxes. As a guest editor I would like to express my sincere appreciation to the contributing authors for their prompt submission of their manuscripts for this issue. All co-authors of the present collection and several QSAR scientists worldwide would like also to devote the present collection to remember the Cuban Prof. Maykel Perez Gonzalez, who was one of the most prolific Cuban authors on QSAR applications to Medicinal Chemistry and promoted the interest of new students in the field. Then, we hope that this issue will not only offer useful and interesting information to scientists who are involved in the field, but, perhaps more importantly, will also serve as an inspiration for new researchers. Finally, on behalf of all co-authors, I extend many thanks to Dr. Allen Reitz, Editor-in-Chief of Current Topics in Medicinal Chemistry for kind attention and help.

The search for new compounds with a given biological activity requires enormous effort in terms of manpower and cost. This effort arises from the large number of compounds that need to be synthesized and subsequently biologically evaluated. For this reason the pharmaceutical industry has shown great interest in theoretical methods that enable the rational design of pharmaceutical agents. In the last years bioinformatics has experienced a great evolution due to the development of specialized software and to the increasing computer power. The codification of the structural information of molecules through molecular descriptors and the subsequent data analysis allow establishing QSAR models (Quantitative Structure-Activity Relationship) that can be applied to the design and the virtual screening of new drugs. The development of sophisticated Docking methodologies also allows a more accurate predict of the biological activity of molecules. Moreover, through this type of computational techniques and theoretical approaches, it is possible to develop explanatory hypothesis on the mechanism of action of drugs. This work provides a brief description of a series of studies implemented in the software MOE (Molecular Operating Environment) with particular attention to the medicinal chemistry aspects.

As the development of the Human Genome Project (HGP), the sequencing of whole human genome has been completed, and a series of human genes have been detected, both of which result in the naissance of pharmacogenomics. Pharmacogenomics is the study of how an individual's genetic inheritance affects the body's response to drugs using the information of human genomics and bioinformatics approaches. It is not only propitious to the rational use of drugs, but also in favor for the personalized drug design.

Artificial neural networks (ANNs) have been widely used for medicinal chemistry modeling. In the last two decades, too many reports used MATLAB environment as an adequate platform for programming ANNs. Some of these reports comprise a variety of applications intended to quantitatively or qualitatively describe structure-activity relationships. A powerful tool is obtained when there are combined Bayesian-regularized neural networks (BRANNs) and genetic algorithm (GA): Bayesian-regularized genetic neural networks (BRGNNs). BRGNNs can model complicated relationships between explanatory variables and dependent variables. Thus, this methodology is regarded as useful tool for QSAR analysis. In order to demonstrate the use of BRGNNs, we developed a reliable method for predicting the antagonistic activity of 5-amino-3-arylisoxazole derivatives against Human Platelet Thrombin Receptor (PAR-1), using classical 3D-QSAR methodologies: Comparative Molecular Field Analysis (CoMFA) and Comparative Molecular Similarity Indices Analysis (CoMSIA). In addition, 3D vectors generated from the molecular structures were correlated with antagonistic activities by multivariate linear regression (MLR) and Bayesian-regularized neural networks (BRGNNs). All models were trained with 34 compounds, after which they were evaluated for predictive ability with additional 6 compounds. CoMFA and CoMSIA were unable to describe this structure-activity relationship, while BRGNN methodology brings the best results according to validation statistics.

Variable selection is a procedure used to select the most important features to obtain as much information as possible from a reduced amount of features. The selection stage is crucial. The subsequent design of a quantitative structure-activity relationship (QSAR) model (regression or discriminant) would lead to poor performance if little significant features are selected. In drug design modern era, by the means of combinatorial chemistry and high throughput screening, an unprecedented amount of experimental information has been generated. In addition, many molecular descriptors have been defined in the last two decays. All this information can be analyzed by QSAR techniques using adequate statistical procedures. These techniques and procedures should be fast, automated, and applicable to large data sets of structurally diverse compounds. For that reason, the identification of the best one seems to be a very difficult task in view of the large variable selection techniques existing nowadays. The intention of this review is to summarize some of the present knowledge concerning to variable selection methods applied to some well-known statistical techniques such as linear regression, PLS, kNN, Artificial Neural Networks, etc, with the aim to disseminate the advances of this important stage of the QSAR building model.

In order to minimize expensive drug failures, is essential to determine potential activity, toxicity and ADME problems as early as possible. In view of the large libraries of compounds now being handled by combinatorial chemistry and high-throughput screening, identification of potential drug is advisable even before synthesis using computational techniques such as QSAR modeling. A great number of in silico approaches to activity/toxicity prediction have been described in the literature, using molecular 0D, 1D, 2D and 3D descriptors. Also these descriptors have been implemented in available computational tools such as DRAGON, SYBYL and CODESSA for it easy use. However, many of them only have been used to explain a few prediction problems. This review attempts to summarize present knowledge related to the computational biological activity prediction based in 2D molecular descriptors implemented in the DRAGON software. These applications rely on new computational techniques such as virtual combinatorial synthesis, virtual computational screening or inverse. Several topological molecular descriptors applications are described, ranging from simple topological indices to topological indices derived from matrices weighted with atomic and bond properties. Their advantages, limitations and its possibilities in drug design are also discussed.

Good progress has been made to modernize traditional Chinese medicines by obtaining active components from natural herbs. In this review, some recent works on procuring active components and modernizing traditional Chinese medicines will be covered. In addition, some recent works on drug design using modern drug design tools have been described. With some well defined targets, the traditional Chinese medicine databases have been screened so as to identify those compounds for which the potential as a drug candidate was not known before. Among these studies, two have been selected as examples to be discussed in details. First, new anti-HIV candidates have been detected, namely leucovorin and agaritine derivatives. Subsequently, GTS-21 is proved to be a good candidate for Alzheimer's disease. All these findings may provide useful information for finding effective drug candidates with lower cost.

In recent times, there has been an increased use of software and computational models in Medicinal Chemistry, both for the prediction of effects such as drug-target interactions, as well as for the development of (Quantitative) Structure-Activity Relationships ((Q)SAR). Whilst the ultimate goal of Medicinal Chemistry research is for the discovery of new drug candidates, a secondary yet important outcome that results is in the creation of new computational tools. The adoption of computational tools by medicinal chemists is sadly, and all too often accompanied, by a lack of understanding of the legal aspects related to software and model use, that is, the copyright protection of new medicinal chemistry software and software-mediated discovered products. This article aims to provide a reference to the various legal avenues that are available for the protection of software, and the acceptance and legal treatment of scientific results and techniques derived from such software. An overview of relevant international tax issues is also presented. We have considered cases of patents protecting software, models, and/or new compounds discovered using methods such as molecular modeling or QSAR. This paper has been written and compiled by the authors as a review of current topics and trends on the legal issues in certain fields of Medicinal Chemistry and as such is not intended to be exhaustive.

The method MARCH-INSIDE (MARkovian CHemicals IN SIlico DEsign) is a simple but efficient computational approach to the study of Quantitative Structure-Activity Relationships (QSAR) in Medicinal Chemistry. The method uses the theory of Markov Chains to generate parameters that numerically describe the chemical structure of drugs and drug targets. This approach generates two principal types of parameters Stochastic Topological Indices (sto-TIs) and stochastic 3D-Topographic Indices (sto-TPGIs). The use of these parameters allows the rapid collection, annotation, retrieval, comparison and mining of molecular and macromolecular chemical structures within large databases. In the work described here, we review and comment on the several applications of MARCH-INSIDE to the Medicinal Chemistry of Antimicrobial agents as well as their molecular targets. First we revised the use of classic sto-TIs to predict antiparasite compounds for the treatment of Fascioliasis. Next, we revised the use of chiral sto-TIs (sto-CTIs) to predict new antibacterial, antiviral and anti-coccidial compounds. After that, we review multi-target sto-TIs (mt-sto-TIs), which unifying QSAR models predicting antifungal, antibacterial, or anti-parasite drugs with multiple targets (microbial species). We also discussed the uses of mt-sto-TIs to assemble drug-drug similarity Complex Networks of antimicrobial compounds based on molecular structure. Last, we review the use of MARCH-INSIDE to generate macromolecular TIs and TPGIs for proteins or RNA targets for antimicrobial drugs.

The drug discovery and development process is lengthy and expensive, and bringing a drug to market may take up to 18 years and may cost up to 2 billion $US. The extensive use of computer-assisted drug design techniques may considerably increase the chances of finding valuable drug candidates, thus decreasing the drug discovery time and costs. The most important computational approach is represented by structure-activity relationships that can discriminate between sets of chemicals that are active/inactive towards a certain biological receptor. An adverse effect of some cationic amphiphilic drugs is phospholipidosis that manifests as an intracellular accumulation of phospholipids and formation of concentric lamellar bodies. Here we present structure-activity relationships (SAR) computed with a wide variety of machine learning algorithms trained to identify drugs that have phospholipidosis inducing potential. All SAR models are developed with the machine learning software Weka, and include both classical algorithms, such as k-nearest neighbors and decision trees, as well as recently introduced methods, such as support vector machines and artificial immune systems. The best predictions are obtained with support vector machines, followed by perceptron artificial neural network, logistic regression, and k-nearest neighbors.

Full text available