Content uploaded by Yvonne C Martin
Author content
All content in this area was uploaded by Yvonne C Martin
Content may be subject to copyright.
A fast new approach to pharmacophore mapping and its application to dopaminergic and benzodiazepine agonist
Abstract
In the absence of a 3D structure of the target biomolecule, to propose the 3D requirements for a small molecule to exhibit a particular bioactivity, one must supply both a bioactive conformation and a superposition rule for every active compound. Our strategy identifies both simultaneously. We first generate and optimize all low-energy conformations by any suitable method. For each conformation we then use ALADDIN to calculate the location of points to be considered as part of the superposition. These points include atoms in the molecule and projections from the molecule to hydrogen-bond donors and acceptors or charged groups in the binding site. These positions and the relative energy of each conformation are the input to our new program DISCO. It uses a clique-detection method to find superpositions that contain at least one conformation of each molecule and user-specified numbers of point types and chirality. DISCO is fast; for example, it takes about 1 min CPU to propose pharmacophores from 21 conformations of seven molecules. We typically run DISCO several times to compare alternative pharmacophore maps. For D2 dopamine agonists DISCO shows that the newer 2-aminothiazoles fit the traditional pharmacophore. Using site points correctly identifies the bioactive enantiomers of indoles to compare with catechols whereas using only ligand points leads to selecting the inactive enantiomer for the pharmacophore map. In addition, DISCO reproduces pharmacophore maps of benzodiazepines in the literature and proposes subtle improvements. Our experience suggests that clique-detection methods will find many applications in computational chemistry and computer-assisted molecular design.
... ZDOCK (29) http://zlab.umassmed.edu/zdock/ Pharmacophore modeling Catalyst (30) -COMSIA (31) -DISCO (32) -GALAHAD (33) -GASP (34) -HipHop (30) -HypoGen (35) -LigandScout (36) https://www.inteligand.com/ligandscout/ MOE (14) https://www.chemcomp.com/Products.htm ...
Virtual screening (VS) techniques have emerged in the past decade as an efficient strategy in lead identification. Molecular docking as well as pharmacophore and 3D-QSAR modelling are two major VS techniques standing out as cornerstones in the process of new drug discovery. Explanation of the main virtual screening techniques as well as the most commonly used validation parameters are discussed thoroughly emphasizing their use and shortcomings. Criteria for the selection of benchmarking datasets and training sets for molecular docking, pharmacophore, and 3D-QSAR models are discussed. Understanding the basics behind these techniques and their validation is crucial to judge the validity of the obtained results. Computational technologies have witnessed great improvement in the last few years which had a great impact on the improvement of virtual screening. Fields such as cloud computing and deep learning algorithms are among the technologies modeling the future of computer-aided drug design. This mini-review gives summarized knowledge for guiding beginners as well as experts in the field of virtual screening.
... This is free software written in Java language that computes different drugrelevant properties to form an active valid structure. It also helps to draw the structure, and calculates drug-relevant properties such as cLogP prediction, solubility prediction, and overall drug-likeness score (Martin et al., 1993). ...
... This is free software written in Java language that computes different drugrelevant properties to form an active valid structure. It also helps to draw the structure, and calculates drug-relevant properties such as cLogP prediction, solubility prediction, and overall drug-likeness score (Martin et al., 1993). ...
Bioinformatics and chemoinformatics are rapidly evolving fields that demand sophisticated algorithms and methods to handle complex data such as genomics data, medical imaging, and electronic medical records. These data can be categorized as big data. During the past few decades, computational methods have impacted biomedical research dramatically. Whether it is an initial screening and prediction of a potential drug candidate or finding a cancer-related pathway, the computational method has proven very useful. Therefore bioinformatics algorithms have been adopted widely as an effective alternative to the laborious time-consuming classical methods in biological research. However, the desired technical and programming experience needed to apply these methods is a big hurdle for the research community to include these remarkable methods in their research. In the recent past, Python has emerged as a very attractive programming language in the field of bioinformatics and chemoinformatics because of its fast learning curve and easy integration in various data analysis pipelines. The main focus of this chapter is to introduce Python to an audience that is not very familiar with the programming language and also provide a hands-on tutorial for those who want to adopt this powerful language in their chemoinformatics research.
... This is free software written in Java language that computes different drugrelevant properties to form an active valid structure. It also helps to draw the structure, and calculates drug-relevant properties such as cLogP prediction, solubility prediction, and overall drug-likeness score (Martin et al., 1993). ...
This chapter emphasizes the important tools of cheminformatics that are proven to be well organized for pharmaceutical data analysis and applications. Cheminformatics acts as an interface between physics, chemistry, biology, mathematics, biochemistry, statistics, and informatics. Cheminformatics refers to solving chemical and synthetic problems effectively by making use of information tools available in the wide space of the web. Each of the tools described in the chapter has been extensively discussed in many research publications but this work gives a complete overview of the major tools and techniques that are able to aid in the study of many important applications such as drug discovery processes and other biochemical applications. It connects molecular modeling with cheminformatics very well. This chapter discusses briefly almost all aspects of molecular modeling and drug designing that utilize cheminformatics tools, ranging from quantitative structure activity relationship analysis, similarity searching to deriving pharmacophores, and shortlisting datasets to forming combinatorial libraries. Cheminformatics aids in designing reactions and possible synthetic routes, structural elucidation of molecules isolated from various biological, and environmental sources or from reaction pathways.
The superposition of small molecules is a standard technique in molecular modeling and for some more advanced in silico applications of drug discovery a critical prerequisite. The aims of superposing molecules are manifold. An assessment of the 3D similarity, an understanding of the SAR in a compound series, or ultimately an estimate of the likelihood of a compound to be active and selective against a target protein of interest. Considering so many objectives it is not surprising that new superpositioning methods are continuously developed and the overlay problem cannot be considered solved. We present 51 superposition methods with a focus on those published in the 21st century. For 36 methods that are currently available, we briefly describe and compare the respective pose generation and scoring processes. While the modeling community got a wealth of methods at hand, the scientific necessity of rigorous and comparable benchmarking becomes apparent.
This article is categorized under: Data Science > Chemoinformatics
Software > Molecular Modeling
The drug discovery paradigm has been very time-consuming, challenging, and expensive; however, the disease conditions originating from bacteria, virus, protozoa, fungus and other microorganisms are steadily shooting up. For instance, COVID-19 is the latest viral infection that affects millions of people and the world’s economy very severely. Therefore, the quest for discovery of novel and potent drug compounds against deadly pathogens is crucial at the moment. Despite a lot of drawbacks in drug discovery and development and its pertaining technology, the advancement must be taken into account so the time duration and cost would be minimized. In this chapter, basic principles in drug design and discovery have been discussed together with advances in drug development.
Melatonin is the principal hormone secreted by the pineal gland, produced in humans with a circadian rhythm characterized by elevated blood levels during the night. It is involved in the regulation of several rhythmic functions in various vertebrates, and participates in the processing of photoperiodic information. Although its role in human physiologic and pathologic processes is not yet completely understood, MLT exerts a number of actions, in physiological or pharmacological concentrations, which could be of interest for future therapeutic uses. The mechanisms involved in MLT actions include interaction with membrane receptors, recently classified as mt 1 /MT 2 /MT 3 , and with nuclear sites corresponding to orphan members of the nuclear receptor superfamily, RZR/ROR; MLT also acts as a radical scavenger, exerting a protective action against various oxidative injuries. The present review is mainly addressed to the medicinal chemistry of ligands at the MLT membrane receptors, focusing on the models of binding interaction published in the literature. Several different pharmacophore and 30-QSAR models have been reported so far, and a re consideration of known active compounds, in the light of the recently developed biological tests on cloned receptors, could help to resolve the incongruities among these models; to this end, additional information is becoming available from new, conformationally constrained ligands, and from antagonist compounds with a selective affinity for receptor subtypes.
Objective
The vast majority of known proteins have not been experimentally tested even at the level of measuring their expression, and the function of many proteins remains unknown. In order to decipher protein function and examine functional associations, we developed "Cliquely", a software tool based on the exploration of co-occurrence patterns.
Computational model
Using a set of more than 23 million proteins divided into 404,947 orthologous clusters, we explored the co-occurrence graph of 4,742 fully sequenced genomes from the three domains of life. Edge weights in this graph represent co-occurrence probabilities. We use the Bron–Kerbosch algorithm to detect maximal cliques in this graph, fully-connected subgraphs that represent meaningful biological networks from different functional categories.
Main results
We demonstrate that Cliquely can successfully identify known networks from various pathways, including nitrogen fixation, glycolysis, methanogenesis, mevalonate and ribosome proteins. Identifying the virulence-associated type III secretion system (T3SS) network, Cliquely also added 13 previously uncharacterized novel proteins to the T3SS network, demonstrating the strength of this approach. Cliquely is freely available and open source. Users can employ the tool to explore co-occurrence networks using a protein of interest and a customizable level of stringency, either for the entire dataset or for a one of the three domains—Archaea, Bacteria, or Eukarya.
Molecular recognition plays a central role in biological systems and is the basis of the specificity seen in antigen-antibody, substrate-enzyme, hormone-receptor, and drug-receptor interactions. The importance of this problem had been recognized by scientists like Pasteur and Ehrlich at the end of the nineteenth century. Structure-activity relations in which one probes the basis of recognition by chemically modifying one of the partners form a significant segment of medicinal chemistry. Attempts to correlate activity with physical properties such as solubility, refractive index, etc. are a reasonable approach to defining the characteristic properties associated with activity in a given biological assay (1). In most systems, these correlations reflect the effective concentration of the drug in the critical biophase rather than the critical features necessary for bimolecular recognition. We would like to concentrate on what can be deduced regarding bimolecular recognition from structure-activity data. A major dichotomy between quantitative structure-activity relations (QSAR) and
Superposition of molecules on a three-dimensional computer graphic display is an efficient means to compare three-dimensional molecular structures. Biologically active molecules, which are presumed to bind to the same receptor site, are thought to have common structural features. But, it is the physical and chemical properties arranged spatially through chemical structures that are important for specific binding to a receptor. Therefore, for the purpose of studies on biological activities, molecules should be superposed to those properties, not to the atomic positions as in the traditional methods. We have developed a program system for realizing this new concept. The concept stands on the general perceptions of organic chemists about hydrogen bondings and chemical isosterisms. The 'goodness of fit' values, which are estimated in realtime on the basis of spatial similarity of those properties between molecules, are displayed and updated throughout the superposing process. This program can construct a receptor cavity model and provide with the cavity size and shape, surface electrostatic potentials, hydrogen bonding sites and so on, by using all information supplied by the superposed molecules. This model can be modified by further superposing of another molecule. These constructed models would be of help for rational drug design, when the receptor structures are not yet known.
Computer-based lead finding algorithms which attempt to design, on a de novo basis, ligands that will complement a known receptor site cavity face some major problems in terms of a combinatorial design space and the synthesizability of the designed molecules. On the other hand, typical 3D database search methods provide a different set of challenges. Both of these approaches are ultimately pointed toward the same goal and can be used together productively. In this article we describe advances in both areas: we first describe extensions to our de novo ligand design software which combines (a) a tree-based conformational search over a library of fragments, and (b) a form of simulated annealing which allows designed ligands to crawl around the binding site even as their structures are changing. In the second part, we then discuss an implementation of the database approach which allows users to formulate 3D substructure, superstructure, or similarity queries based upon demonstrated or hypothetical requirements for activity. Finally, we draw the two approaches together with an example of current research interest, showing how one method can feed the other.
This paper illustrates a method of searching for matched accessible surfaces on two dissimilar molecules without specifying a particular face to be matched. Both molecules are allowed to rotate and their residuals are minimized to obtain matches. To identify matched orientations, 6-dimensional cluster analysis is used.
This paper provides a method for searching for pattern matches between parameters mapped onto molecular surfaces. A window, of any specified shape, is projected on to one molecule and a second molecule is rotated to find discrete matches within the projected windows. This search is optimized at two levels. Level 1 finds feasible regions for matching. Level 2 ascertains the best match in a feasible region through an intermediate application of cluster analysis to level 1 data; this is followed by further extensive minimization to locate distinct matching orientations. The method has been tested for the accessible surface and the electrostatic potential surrounding the neurotoxins saxitoxin and tetrodotoxin.
A series of rigid tricyclic analogues of the dopamine (DA) agonist PD 118440 [4-(1,2,5,6-tetrahydro-1-propyl-3-pyridinyl)-2-thiazolamine] was synthesized and evaluated for dopaminergic activity and DA autoreceptor selectivity. (R)-(+)-6-Propyl-4,5,5a,6,7,8-hexahydrothiazolo[4,5-f]quinolin-2-amine ((+)-6) was identified as the most selective DA autoreceptor agonist from this group of compounds. It inhibited spontaneous locomotor activity (LMA) in rodents, reversed the gamma-butyrolactone (GBL) induced accumulation of rat striatal DOPA and inhibited brain DA neuronal firing, all suggestive of direct DA autoreceptor agonist activity. However, (+)-6 is not completely free of postsynaptic DA activity, as evidenced by its stimulation of LMA in rats at high doses and its ability to produce stereotypy. On the other hand, (-)-6 appears to be a weak partial DA agonist with some effects on brain DA synthesis only at high doses. Like other DA autoreceptor agonists and DA antagonists, (+)-6 inhibited Sidman conditioned avoidance in squirrel monkeys, a test predictive of clinical antipsychotic activity. However, unlike classical antipsychotics, (+)-6 did not induce dystonias in haloperidol-sensitized squirrel monkeys, suggesting a minimal propensity toward extrapyramidal side effects (EPS).
This paper outlines a method for gnomonic projection of a molecular surface and a novel application of it to the problem of surface comparison. Semiregular arrays of points are generated by icosahedral tessellation. The surface may be the accessible surface or a chemical parameter surface such as the molecular electrostatic potential. Gnomonic projection retains the 3D characteristics of the inspection surface. Comparison of two surfaces can be achieved by statistical assessment of the pattern match. The method opens the gateway to an optimized search for pattern matches on the surfaces of dissimilar molecular structures.
In order to analyze the directionality of hydrogen bonding to oxygen atoms the Cambridge Crystallographic Data File was searched for O. . . X (X equals N, O) intermolecular contacts at less than 3 A in structures containing ether, ketone, epoxide, enone, and ester groups. The results are represented as scatterplots. However, for further clarity, they are also represented as diffuse (probability) densities obtained by superposing spherical atomic electron density functions with a parameterizable temperature factor on each point in the scatterplot and contouring the maps at a height proportional to the number of data points. In all systems the largest concentration of hydrogen-bonded X groups lay in the direction commonly ascribed to lone pairs, but generally only one such X group per oxygen atom.
Geometrical analysis of hydrogen bonds observed in crystal structure data retrieved from the Cambridge Crystallographic Data File reveals lone-pair directionality and leads to an extended hydrogen bond potential function for use in molecular mechanics calculations. This potential function includes as variables the hydrogen-acceptor separation, the angle subtended at the hydrogen atom, the angle at the acceptor atom, and the displacement of the hydrogen atom from a defined plane containing the lone-pair orbitals of the acceptor atom. With such directional terms included in the force field, electrostatic contributions to the force-field energy can be reduced or even entirely eliminated without adverse consequences, thus removing many problems associated with the assignment of atomic partial charges and an effective dielectric constant. The new hydrogen bond potential function has been incorporated in the molecular mechanics program YETI and used to refine the conformation of four complexes of human carbonic anhydrase II with small molecules: the natural substrate bicarbonate and three heterocyclic sulfonamide inhibitors.
