Year : 2009 | Volume
: 1 | Issue : 1 | Page : 16--22
Aarti Sharma1, Himanshu Gupta2,
1 School of Pharmacy, Jaipur National University, Jagatpura, Jaipur, India
2 Faculty of Pharmacy, Jamia Hamdard (Hamdard University), New Delhi, India
Faculty of Pharmacy, Jamia Hamdard (Hamdard University), New Delhi
The use of computational chemistry in the development of novel pharmaceuticals is becoming an increasingly important tool. In the past, drugs were simply screened for effectiveness. The recent advances in computing power and the exponential growth of the knowledge of protein structures have made it possible for organic compounds to be tailored to decrease the harmful side effects and increase the potency. This article provides a detailed description of the techniques employed in molecular modeling. Molecular modeling is a rapidly developing discipline, and has been supported by the dramatic improvements in computer hardware and software in recent years.
|How to cite this article:|
Sharma A, Gupta H. Molecular modeling.J Pharm Bioall Sci 2009;1:16-22
|How to cite this URL:|
Sharma A, Gupta H. Molecular modeling. J Pharm Bioall Sci [serial online] 2009 [cited 2019 Oct 16 ];1:16-22
Available from: http://www.jpbsonline.org/text.asp?2009/1/1/16/62681
Drug design is an iterative process, which begins when a chemist identifies a compound that displays an interesting biological profile and ends when both the activity profile and the chemical synthesis of the new chemical entity are optimized. Traditional approaches to drug discovery rely on a step-wise synthesis and screening program for a large number of compounds, to optimize the activity profiles. Over the past ten to twenty years, scientists have used computer models of new chemical entities to help define activity profiles, geometries, and reactivities.
Molecular Modeling is one of the fastest growing fields in science. It may vary from building and visualizing molecules.  In molecular modeling we encounter all types of molecules such as Structural (graphical, ball and stick, wire), Phenomenological (Homology, secondary structure prediction), and Mathematical (computer simulations). The development of computerized molecular modeling (CMM) had made traditional models less favorable in the late 1960s. Not only are computers capable of drawing and manipulating molecules in three dimensions, they are also powerful tools for predicting molecular spatial structure through energy minimization calculations, based on quantum mechanics.
Computational Molecular Models
Computational molecular models are the result of mathematical equations that estimate the positions and responses of electrons and nuclei. The mathematical models are divided into classical mechanical and quantum mechanical approaches. Classical mechanics looks at molecules as a collection of atoms and bonds that are treated as balls and springs. Information like atomic radii and spring stiffness are used to find the 'best' position of atoms. It is a fast and accurate method for locating a molecule's stable geometry. Quantum mechanical methods solve Schrodinger's equation in two ways: semi-empirically and ab initio (from the beginning). Semi-empirical methods use experimental data to simplify the solution of the Schrodinger's equation, so it can be solved faster. Many methods for this simplification have been developed, including Huckel, Extended Huckel, INDO/S, and MNDO. Each has a parameter set based on experimental measurements for numerous compounds. In contrast, ab initio methods use only standard mathematical approximations. While these methods are theoretically 'pure,' they are computationally intensive and rely on large, ultra high-speed computers.
Molecular Modeling is the study of molecules and molecular systems by means of computer models. The most common model for a molecule consists of a set of atoms, considered to be a point mass having a position in space and possibly other properties. All the atoms in a molecular system interact with each other; these interactions are described by the interaction energy, which is a function of the positions and properties of the atoms. The interaction is the central quantity in molecular modeling, and constructing an appropriate function is one of the major problems. There are many 'recipes' for calculating the interaction energy, ranging from very simple functions of interatomic distances to the solution of the Schrodinger equation (the fundamental equation of quantum mechanics), for all electrons of all atoms in the system. The recipes are called force fields.
Construction and manipulation of these molecules use Biology (for function, sequence, genetics, and modifications), Chemistry (for the structure of the component parts and the knowledge of what holds them together), Physics (for concepts of Energy, Forces, Statistical Mechanics, and Thermodynamics), and Computer Science (to represent, manipulate, visualize, and organize models).
To use molecular modeling in a rigorous fashion it is necessary to understand the relationship of the 3D structure of a system to its energy. Various mathematical models relate the structure to energy.
Various Aspects of Molecular Modeling
Molecular orbital theory / Linear combination of atomic orbitals. Orbitals (electrons) treated explicity. High accuracy but limited to 100 atoms or less.
Molecular Mechanics (MM) is a computational technique used to model the conformational behavior and energetic properties of molecules. The molecule is treated at the atomic level, that is, the electrons are not treated explicitly. MM uses an Energy Function defined so that given a particular conformation, (i.e., given a set of spatial coordinates for all the atoms) the energy of the molecule can be calculated. The energy function is empirical, that is, it is not entirely derived from rigorous theories. Usually a combination of quantum mechanical calculations and experimental data are used to construct the energy function. 
The energy function makes a distinction between 'bonded' and 'non-bonded' interactions. Bonded interactions occur between atoms that are connected by no more than three bonds. Non-bonded interactions occur between atoms that are either not connected at all (e.g., atoms on two residues that are separated, far from each other, in a polypeptide chain) or are only connected by more than two bonds. (e.g., the first and last carbon atoms of n-pentane).
Bonding is the interaction between two atoms directly bonded to each other, and is assumed to be harmonic (imagine a spring joining the two atoms - it prevents the atoms getting too close and too far away). Bond Angles is the interaction between three connected atoms, which is also assumed to be harmonic (imagine a spring that maintains the angle between the two bonds at some value, e.g., 109. for a H-C-H angle in methane). Dihedral (Torsion) Angles is found in the case of four atoms connected by three bonds, and we look straight down the second bond, the dihedral angle is the angle between the first and third bonds. The energies associated with dihedral angles are treated using a cosine series: this allows the angle to have several preferred values. For example, in ethane, the H-C-C-H dihedral angle prefers to be at 60°, 180° or 240° (i.e., - 60°).
Van der Waals and repulsive interactions operate between all atoms. They attract at long range (due to London Dispersion forces), but are strongly repulsive at a short range (because atoms cannot overlap). Coulombic interactions operate between all atoms that carry a charge. Obviously Na + carries a charge of +1e. CH 3 OH (methanol) carries an overall charge of 0e. However, oxygen is a much more electronegative element than either carbon or hydrogen, so it is in effect slightly negatively charged, and would therefore be assigned a partial charge of say -0.5e, with maybe +0.4e being assigned to the hydroxyl hydrogen and +0.1e being assigned to the carbon.
The energy function
E = E bonds + E angle + E dihedral + E non − bonded
This function, referred to as a potential function, computes the molecular potential energy as a sum of energy terms that describe the deviation of bond lengths, bond angles, and torsion angles away from the equilibrium values, plus terms for non-bonded pairs of atoms describing van der Waals and electrostatic interactions. The set of parameters consisting of equilibrium bond lengths, bond angles, partial charge values, force constants, and van der Waals parameters are collectively known as the force field. Different implementations of molecular mechanics use slightly different mathematical expressions, and therefore, different constants for the potential function. The various force fields commonly used in biomolecular simulations are AMBER, GROMOS, and CHARMM specifically designed to describe the most common biological units such as amino acids and nucleic acids. 
Energy minimization (EM) techniques are used to iteratively refine the conformation of a molecule so as to minimize its energy. EM is particularly useful for removing bad clashes between atoms that may have developed when building or modeling a molecule (e.g., a homology-modeled structure of a protein). There are many methods that can be used - all have their own strengths and weaknesses - Steepest Descent (SD), Conjugate Gradient (CG), Second Order Methods. All EM techniques concentrate on searching downhill - they therefore tend to find the nearest local minimum on the energy surface. If a much deeper (i.e., better) energy minimum is nearby, but separated from the starting point by a high energy barrier, it will not be found. Energy minimization is therefore not capable of finding the global energy minimum - the absolute best conformation of a molecule will only be found if we start very close to it. 
Molecular dynamics combines energy calculations from the force field methodology with the laws of Newtonian (as opposed to quantum) mechanics. Molecular dynamics (MD) simulations are used to simulate the dynamic behavior of molecular systems. Starting with a suitable initial conformation of our system, we used Newton's equations of motion over small time steps (usually 10 -15 sec or 1 fsec) to determine how the system evolves over time. The simulation is initialized by providing the location and assigning a force vector for each atom in the molecule. The acceleration of each atom is then calculated from the equation a = F/m, where m is the mass of the atom and F the negative gradient of the potential energy function (the mathematical description of the potential energy surface). A long MD simulation of a protein for example, provides insight into its conformational flexibility. The simulation can tell you what conformational states are accessible to the protein, and the timescale of these conformational fluctuations. The Verlet algorithm is used to compute the velocities of the atoms from the forces and atom locations. Once the velocities are computed, new atom locations and the temperature of the assembly can be calculated. These values are then used to calculate trajectories or time-dependent locations for each atom. The trajectory generated from an MD simulation can also be used to calculate the thermodynamic properties (e.g., heat capacity). 
Molecular dynamics simulations have been used in a variety of biomolecular applications. The technique, when combined with data derived from Nuclear Magnetic Resonance (NMR) studies, has been used to derive 3D structures for peptides and small proteins in cases where X-ray crystallography is not practical.  Additionally, structural, dynamic, and thermodynamic data from molecular dynamics has provided insights into the structure-function relationships, binding affinities, mobility, and stability of proteins, nucleic acids, and other macromolecules that cannot be obtained from static models.
The key advantage of MC and MD computer simulation methods over various experimental techniques lies in their ability to allow scientists to describe in exquisite, atomistic detail, the behavior of individual molecules. These methods are also able to track the time evolution of the structure and interactions of molecules from the femto to the microsecond scale. This level of sophistication enables the analysis of nearly every conceivable property of single molecules or bulk materials.
Electrostatic interactions are very long-ranged (recall the 1/r dependence of the Coulombic term in the MM energy function). They can accelerate the rates at which molecules associate, for example, acetylcholinesterase. Some of the more common biological macromolecules are highly charged so we cannot simply ignore electrostatic interactions.  One of the fundamental equations of classical electrostatics is the Poisson-Boltzmann equation. The calculations discussed so far assumed the molecules to be completely rigid. In other words, one used only one conformation of a molecule or a molecular complex in a calculation. However, many molecules are flexible in a solution, at physiological temperature. One way to deal with the flexibility issue is to carry out molecular dynamics simulations, to generate an ensemble of structures for continuum electrostatic calculations. Some recent works have been done along this line and encouraging results have been obtained. However, due to the assumptions employed in these models, more work is required to further validate these methods. 
Construction of the molecular model
The molecular model is built by fitting a known sequence into an approximate electron density, derived from some phasing model. This process is subject to errors, which are corrected as the refinement continues. The model is usually built with good local geometry, but global geometry issues are cleared by refinement. During model building the following errors can occur. 
The right atom in the wrong density. Density can be misinterpreted so the wrong piece of structure is built into it. This can be particularly common with respect to placement of solvent and surface side chains. Atoms with no supporting density. This can occur for a variety of reasons, most common of which is disorder.Atoms appear where no atoms should occur. Noise occurs in three dimensions, and noise peaks can look like atoms.Chain registration errors. The electron density carries no labels, so it is easy to be off by a few residues when following the chain. This is particularly likely after tracing through a surface loop or another area of weak density.Mistracing of the main chain. Frequently the chain is traced in several pieces, which must then be connected. It is possible to misconnect the pieces. It is also possible to trace a piece of chain backwards.Applications of molecular modeling
Molecular modeling of dendrimers for nanoscale applications
Dendrimers are well defined, highly branched macromolecules that radiate from a central core and are synthesized through a stepwise, repetitive reaction sequence that guarantees complete shells for each generation, leading to polymers that are monodisperse. The synthetic procedures developed for dendrimer preparation permit nearly complete control over the critical molecular design parameters, such as size, shape, surface / interior chemistry, flexibility, and topology. Recent results suggest that dendritic polymers may provide the key to developing a reliable and economical fabrication and manufacturing route to functional nanoscale materials, which would have unique properties (electronic, optical, opto-electronic, magnetic, chemical or biological). In turn, these could be used in designing new nanoscale devices.  For example poly(amidoamine) (PAMAM) dendrimers have been used, to attract copper (II) ions inside the macromolecules, where they are subsequently reacted with solubilized H2S to form metal sulfides.  These organic / inorganic, dendrimer-based hybrid species have been termed 'nanocomposites' and display unusual properties. For example, the solubility of the nanocomposites is determined by the properties of the host dendrimer molecules. This allows for solubilization of the inorganic guest compounds, in environments where they are inherently insoluble. As it has been established that there is no covalent bond between host and guest, these observations suggest that the inorganics are physically and spatially restricted by the dendrimer shell. However, this structure has not been verified. Here we use 3D structure building tools based on the CCBB MC method and molecular dynamics techniques, to investigate the structural characteristics of PAMAM dendrimers. 
Pharmacophore based docking
In the field of molecular modeling, docking is a method that predicts the preferred orientation of one molecule to a second, when bound to each other, to form a stable complex. Knowledge of the preferred orientation, in turn, may be used to predict the strength of the association or binding affinity between the two molecules by using, for example, the scoring functions.
The associations between biologically relevant molecules such as proteins, nucleic acids, carbohydrates, and lipids play a central role in signal transduction. Furthermore, the relative orientation of the two interacting partners may affect the type of signal produced (e.g., agonism vs. antagonism). Therefore docking is useful for predicting both the strength and type of signal produced. Docking is frequently used to predict the binding orientation of small molecule drug candidates to their protein targets, in order to, in turn, predict the affinity and activity of the small molecule. Hence, docking plays an important role in the rational design of drugs. Given the biological and pharmaceutical significance of molecular docking, considerable efforts have been directed toward improving the methods used to predict docking. 
Computer-aided retrometabolic drug design: Soft drugs
Soft drug design approaches aim to design new therapeutic agents that undergo facile, preferentially hydrolytic metabolism to produce inactive metabolites. This approach is general and can be used in essentially any therapeutic area, especially where the desired activity is localized, relatively short-lived or susceptible to easy titration. In most cases, this approach aims to design close steric and electronic analogs of an existing drug, which serves as a lead for the design. Therefore, computer programs that can generate virtual libraries of possible analogs and can provide quantitative tools to rank them on the basis of the closeness of their properties to the original lead, are of particular relevance. 
Determination of drug excipient interactions
The molecular modeling technique became popular to study the drug-excipient interaction, which helps to visualize the type and site of interaction on a computer monitor. It was reported in a study that seven glucose units were combined to get a well shaped energy minimized conformation. The cavity depth and the diameter of a wider and narrower rim were calculated and compared to the literature values using the DTMM package. Similarly, norfloxacin, ciprofloxacin, and other structures were built to get energy minimized conformation. The dimensions of these molecules were measured and compared to the literature values. The drug molecules were allowed to penetrate through the cavity and the probability of penetration was observed. Finally, the success in the formation of an inclusion complex of betacyclodextrin with norfloxacin, ciprofloxacin, tinidazole, and methotrexate was reported. 
Quantitative structure activity relationship studies
Quantitative structure activity relationship (QSAR) is a technique that quantifies the relationship between structural and biological properties. A QSAR can be expressed in its most general form by the following equation: Biological activity = f (physicochemical and / or structural parameters). The physicochemical descriptors include parameters that account for hydrophobicity, topology, electronic properties, and steric effects, and are determined empirically by computational methods. Activities used in QSAR include chemical measurements and biological assays. Researchers have attempted for many years to develop drugs based on QSAR. An example of QSAR in modeling is a series of 1-(X-phenyl)-3,3-dialkyl triazenes. The compounds are of interest for their anti-tumor activity, but they are also mutagenic. QSAR is applied to understand how the structure may be modified to reduce the mutagenicity, without significantly decreasing the anti-tumor activity. In a quantitative activity relationship study, the antileishmanial activity of the substituted pyrimidine and pyrazolo pyrimidine analogs was determined using physicochemical and steric descriptions (hydrophobicity, molar refractivity, Suptons resonance, Verloop's steric parameters, and van der Waals volumes of the substituent groups) of the varying substituents. The study of pyrimidine analogs indicated the necessity of having unsubstituted pyrimidine for antileishmanial activity. A linear multiple regression analysis with least square method was applied in developing a correlation. 
A lead is any chemical compound which shows biological activity. It is not the same as a drug molecule, but its generation is an important step in the drug discovery process. It is the process of identifying potential drug compounds or leads that interact with a target with sufficient potency and selectivity. Lead generation is a complex process, which involves two basic steps; Lead finding, is a step to find a chemical compound that has a desired biological activity; Lead optimization, involves elaborating around the basic lead structure to build in all the desirable properties, such as, safety, solubility, and so on, with the help of molecular modeling.
Determination of properties of a pharmacophoric pattern
A pharmacophoric pattern may be defined as a geometrically arranged functionality, possessed by a set of active compounds having some mechanism of action. Identification of the pharmacophores is especially useful in designing receptor agonists and antagonists, enzyme inhibitors, and so on. The molecular modeling approach has been particularly rewarding in dopamine agonists, antagonists, and in drugs acting on histamine and morphine receptors. 
Biomolecular modeling: From drug discovery to nanotechnology
Modern pharmaceuticals, food, materials, and nanotechnology industries increasingly rely on the rationalization and prediction of molecular structure, stability, and function, in order to optimize their products, reduce the time and cost of development, and increase their success rate. The methods of computer simulation and molecular modeling, such as, Monte Carlo (MC) and molecular dynamics (MD), continue to come of age, making significant contributions to a wide variety of experimental fields, ranging from molecular biology and drug design to nanotechnology and biomaterials design. Concomitant rapid advances in computational power have enabled these technologies to tackle physical, chemical, and biological molecular phenomena of unprecedented high complexity (e.g., protein-protein interactions), long time scales (e.g., protein folding), and / or long length scales (e.g., polymer-nanoparticle interactions). 
Our research efforts are aimed at investigating the molecular forces that determine the stability and activity of biomolecules, the behavior of polymeric drug delivery systems, and the specificity and strength of drug-protein interactions. Furthermore, we aim to develop computational algorithms and methods that impact drug discovery and delivery, particularly in major diseases such as diabetes, cancer, and Alzheimer's disease. These developments and their outcomes have a great potential for the generation of intellectual property, which is of significant value to the pharmaceutical, biomaterials, nanotechnology, and software industries. The Biomolecular Modeling Group has interests in a number of specific areas.
Protein flexibility and solvation in drug design
The influence of water is critical in determining the specificity, geometry, and affinity of biomolecular interactions. This arises from the entropic effects and desolvation penalties that determine the magnitude of the free energies of binding. In recent years we have been developing methods for incorporating an explicit treatment of hydration into a computer-aided drug design. The next step is to implement a realistic method for the dynamic hydration of ligand-protein complexes within the docking and structure-based drug design applications, which can deal with water-mediated contacts and desolvation. In the case of protein-protein interactions, it is also essential to determine the magnitude of water-mediated interactions on the free energies of binding between two protein molecules or between one of the proteins and a drug molecule. These outcomes are of significant importance to the pharmaceutical industry, as the lack of success in many drug discovery projects can be partly attributed to neglecting the aqueous environment in which biological interactions take place. 
Protein flexibility at both levels of local side chain rearrangements and large inter-domain conformational changes also plays a crucial role in drug-protein and protein-protein interactions of therapeutic interest. In recent times, methods have been developed for predicting and / or incorporating the effects of protein flexibility in drug design. Further research is now required to enable the prediction of protein conformational changes upon binding to a drug or second protein, as is the case of the insulin receptor. Many current therapeutic targets involve proteins that undergo conformational changes as part of their normal function (e.g., insulin receptor, protein kinases, GPCRs), creating a need in the pharmaceutical industry for an appropriate method of characterizing and predicting such conformational changes and the way they affect the design of drugs. The joint treatment of explicit hydration and protein flexibility will allow for the modeling of ligand-induced conformational changes in aqueous solution, creating an accurate approach to the modeling of ligand-protein interactions.
We have recently been involved in the development of new ligand-protein docking optimization methods and the analysis of conformational, hydrogen-bonding and hydrophobic properties of the ligand binding site. We are working on developing a description of the steric properties of binding sites and identifying regions within a binding site of unique geometric and overall interaction properties. This will solve the problem often found in drug design when selecting, prioritizing, and partitioning a large number of hydrogen-bonding, hydrophobic, and steric features of a binding site or protein interface. At the same time, these new methods can be applied to drug discovery efforts in therapeutic areas such as diabetes and cancer. 
Protein-protein interactions involve extended interfaces, which complementarily determine the free energy of the interaction. We have recently been involved in the development of new methods for computing protein-protein interaction-free energies. Methods are developed by looking at selected 'mutants' of protein complexes in order to disentangle their steric, electrostatic, hydrophobic, and water-mediated contributions. Molecular dynamic simulations help determine the role of the dynamic plasticity of amino acid side chains and water molecules in determining the strength and specificity of the interaction. 
Drug loading and delivery in nanoparticles
In the area of drug delivery, biodegradable copolymeric materials have been used for some time to achieve surface erosion for the controlled delivery of embedded drug molecules. The delivery of drug molecules is dependent upon the rate of solvent penetration, pH, and ionic force, responsible for the gradual degradation of the polymeric surface. Dendrimeric materials have also been used to physically encapsulate drug molecules through hydrophobic interactions or steric impediment. However, at present, there is limited knowledge of the molecular mechanisms responsible for the successful loading and delivery of drug molecules. We can use computer simulation methods to provide an insight into the molecular mechanisms underlying the interactions of the drug with the designed chemical components of the nanoparticle and the role of additives in enhancing these interactions. 
One application of polymeric nanoparticles has been the targeting of the blood-brain barrier to enhance the delivery rate and specificity of treatments for analgesia and Alzheimer's disease. Computer simulation methods are being used to investigate the molecular structure of encapsulated drug molecules in copolymeric biodegradable materials, the mechanism of penetration of water and its pH and ionic force dependence, and the mechanism of interaction of the drug with the various chemical polymeric components that might influence its loading and delivery.
The hydrophobic effect
The hydrophobic effect is the archetypal solvent-induced force, arising from the intermolecular ordering processes in water that occur in the vicinity of non-polar species in an aqueous solution. This phenomenon has been widely studied due to its importance in many chemical and biological processes, such as: the solubility of drug molecules, the adsorption of surfactants onto surfaces, the formation of micelles and biomembranes, the interactions between macromolecules and the association of ligands and proteins in a solution, and the folding and stability of proteins. Computer simulation techniques are used to characterize the free energy changes and the solvent structure and dynamics in simple but realistic systems, by modifying the solute concentration, size and curvature, temperature, pressure and / or salt concentration. 
Protein denaturation and stabilization
Chemical reagents such as urea, guanidium chloride, and ethanol exhibit characteristic molecular interactions with proteins during their denaturation. Interestingly, these agents have the ability to increase the solubility of non-polar molecules in water and to increase the critical micelle concentration. It is important for the food and biopharmaceutical industries to be able to predict the conditions that favor the stability of proteins in, for example, foodstuffs and biological reagents. Computer simulations are being used to determine the structure and dynamics of hydration in model systems such as, water / urea and water / alcohol / urea solutions, where there is experimental (structural and thermodynamic) data available. In the case of proteins, computer simulations are being used to determine the nature of the interactions of the protein surface with such denaturant agents, as well as the degree of solvent accessibility at different denaturant concentrations and temperatures when compared to water accessibility. This will help determine the extent of protein chemical degradation under various solvent formulations. 
The mechanism of cryoprotection
Aqueous mixtures of solvents such as DMSO, glycerol, and ethylene glycol are widely used as cryoprotective agents to preserve biological tissues during freezing. It is believed that these agents suppress crystallization in cell water by inducing the formation of a glassy state, preventing hyperosmotic injury of the tissues caused by sodium chloride. An analogous mechanism is said to allow the stabilization of proteins by sugars, at low hydration levels. However, there is the need for molecular modeling and computer simulation studies to assist the explanation of the molecular mechanism of cryoprotection, as some solvents have the above-mentioned properties and others do not. This knowledge would help with the design of better solvent mixtures to improve the cryopreservation properties and cooling / heating rates used.
In recent years a number of computer simulations have been carried out to investigate the effect of temperature and DMSO concentration in aqueous solutions near room temperature. These studies have validated the recently developed intermolecular potentials of DMSO and have shed a new light on the hydrogen-bonding structure that develops in water / DMSO mixtures. 
Methods for identifying compounds useful for producing heavy oils from underground reservoirs
The methods facilitate the development of chemicals with improved physicochemical features. The methods generally involve first identifying a physicochemical property of a compound that needs to be improved, in order to increase the efficiency of heavy oil removal from underground reservoirs. Next, the physicochemical property is calculated by molecular modeling using semi-empirical or ab-initio calculations. By modifying the molecular model of the compound, the targeted physicochemical property can be optimized. After a suitable compound has been identified, the compound can be synthesized and evaluated. 
Different software are available for Drug Design and Molecular Modeling like Sybyl, Alchemy, Amber, Chem. Office, MOE (molecular operating environment), and Cerius as a tool for Computer Aided Drug Design (CADD), Computer Assisted Molecular Design (CAMD), and Computer-Assisted Molecular Modeling (CAMM), which consider 3D aspects of drugs and provid: 
The 3-D structure of the moleculesThe chemical and physical characteristics of the moleculeComparison of the structure of one molecule with other different moleculesVisualization of a complex formed between different moleculesPrediction about how related molecules might lookMolecular mechanicsMolecular dynamicsConformational searchingDocking studies
|1||Available from: http://www.nyu.edu/pages/mathmol/quick_tour.html. [retrieved on 2009 Oct 9].|
|2||Available from: http://www.en.wikipedia.org/wiki/Molecular_modeling. [retrieved on 2009 Oct 9].|
|3||Leach AR, Taylor JB, Triggle DJ, editors. Ligand-based approaches: Core molecular modeling. Vol. 4, Chapter 4.05, In: Comprehensive medicinal chemistry II, Computer-assisted drug design. NY, Elsevier; 2006. p. 87-118.|
|4||Available from: http://www.en.wikipedia.org/wiki/Energy_minimization. [retrieved on 2009 Oct 9].|
|5||Marshall GR, Louis ST, Beusen DD, Abraham DJ, editors. Molecular modeling in drug design. Burger′s medicinal chemistry and drug discovery. Vol.1, 6 th ed. Virginia: John Wiley and Sons, Inc.; 2003. p. 77-168. |
|6||Available from: http://www.netsci.org/Science/Compchem/feature01.html. [retrieved on 2009 Oct 9].|
|7||Available from: http://www.spdbv.vitalit.ch/TheMolecularLevel/ModQual/. [retrieved on 2009 Oct 9].|
|8||Sims M. CEO Nanorex Inc. Bloomfield Hills, Mich. http://www.machinedesign.com/article/molecular-modeling-in-cad-0928. [accessed on 2006 Sep 28].|
|9||Cagin T, Wang G, Martin R, Breen N, Goddard WA. Molecular modeling of dendrimers for nanoscale applications. Nanotechnology 2000;11:77-84.|
|10||Naylor AM, Goddard WA 3 rd , Keiffer GE, Tomalia DA. Starburst dendrimers 5 molecular shape control. J Am Chem Soc 1989;111:2339-41.|
|11||Tomalia DA, Hedstrand DM, Ferritto MS. Combburst dendrimer topology-new macromolecular architecture derived from dendritic grafting. Macromol 1999;24:1435-8.|
|12||Joseph-McCarthy D, Thomas BE 4th, Belmarsh M, Moustakas D, Alvarez JC. Pharmacophore-based molecular docking to account for ligand flexibility. Proteins 2003;51:172-88.|
|13||Computer-aided retrometabolic drug design: Soft drugs. Expert Opin Drug Discov 2007;2:923-33.|
|14||Nadendla RR. Molecular modeling: A powerful tool for drug design and molecular docking. Resonance 2004;9:51-60.|
|15||Available from: http://www.en.wikipedia.org/wiki/Quantitative_structure-activity_relationship. [retrieved on 2009 Oct 9].|
|16||Van Drie JH. Strategies for the determination of pharmacophoric 3D database queries. J Comput Aided Mol Des 1997;11:39-52.|
|17||Available from: http://www.wabri.org.au/research/molec_modeling.html. [retrieved on 2009 Oct 9].|
|18||Carlson HA. Protein flexibility and drug design: How to hit a moving target. Curr Opin Chem Biol 2002;6:447-52.|
|19||Available from: http://www.pubs.acs.org/subscribe/journals/mdd/v04/i05/. [retrieved on 2009 Oct 9].|
|20||Mancera RL. Biomolecular modeling: from drug discovery to nanotechnology. Availa-ble from: http://www.wabri.org.au/research/molec_modeling.html. [retrieved on 2009 Oct 9].|
|21||Horvath SG, Longpre M, Zabel F. Methods for identifying compounds useful for producing heavy oils from underground reservoirs. WO/2009/061555; 2009.|
|22||Joshi JK. Importance of 3D structure of drugs in drug design and drug receptor interac-tions. Pharmainfo Net 2007;5:5.|