Embed Size (px)
Citation preview
PRESENTED BY:
SURMIL SHAH2 N D SEMESTER M.PHARM
SOFTWARE RESOURCES
CONTENT
KEY WORDS MOLECULAR MECHANICS(MM3 &MM4) HUCKEL-MO-CALCULATOR MOPAC (semi-empirical) GAUSSIAN(Ab initio) SciQSAR PDB SYBYL CHARMM HyperChem MM+ AMBER MOE DISCOVERY STUDIO FlexX GOLD SCHRODINGER Chem3D ALCHEMY 2000
KEY WORDS
• CNDO - Complete Neglect of Differential Overlap: spherically symmetric orbitals only
• INDO - Intermediate Neglect of Differential Overlap: one centre repulsion integrals between orbitals on the same atom.
• MINDO - Modified Intermediate Neglect of Differential Overlap: empirical data to parameterize the repulsion integrals rather than analytic solutions
• NDDO - Neglect of Diatomic Differential Overlap: includes directionality of orbitals on the same atom for repulsion integrals
• MNDO - Modified Neglect of Differential Overlap: better determination for multi-centre repulsion integrals
MOLECULAR MECHANICS (MM3 AND MM4)
(A) MM3Bonds, angles and torsions are not
independent in molecules. They couple with one another. Cross terms between all contributions in a
force field should be considered, especially when calculating vibrational frequencies.
Most cross terms involve two internal coordinates such as stretch-stretch, bend-bend, stretch-bend, stretch torsion, etc.
MOLECULAR MECHANICS (MM3 AND MM4)
Some force fields (MM2, MM3) include cross terms involving three or more internal coordinates, for example stretch-stretch-bend.
Based on QM calculations Dinur and Hagler (1991) proposed the most important terms to be: stretch-stretch, stretch-bend, bend-bend, stretch-torsion and bend-bend-torsion.
Popular cross terms expressions
Ustr−str = 1 2 kab,bc(rab − rab,0)(rbc − rbc,0)
Ubend−bend = 1 2 kabc,dbe(θabc − θabc,0)(θdbe − θdbe,0)
Ustr−bend = 1 2 kab,abc(rab − rab,0)(θabc − θabc,0)
Ustr−tors = 1 2 kbc,abcd (rbc − rbc,0)(1 +
cos(nφ − ψ))
CROSS TERMS
Presence of cross terms can characterize force fields.
Class I - harmonic potential with no cross terms. Class II - includes cross terms and and harmonic
terms. Class II force fields aim at reproducing not only geometry but also vibrational frequencies.
Class III - accounts for electronegativity, polarizability or hyperconjugation.
By adding extra terms to a potential energy expression one obtains a more accurate force field. More parameters will better match the experimental data. Parameterization of more parameters is more complicated and time consuming.
MOLECULAR MECHANICS (MM3 AND MM4)
(B) MM4MM4 is the latest in the well-known series of
molecular mechanics programs developed by Prof. N. L. Allinger and co-workers at the University of Georgia.
The high quality geometries and energies generated by these applications have led to their widespread recognition as the industry standard for calculating small molecule structures.
MM4 continues this trend, offering increased accuracy in calculating molecular structures, performing conformational analyses, and computing spectroscopic and thermodynamic properties.
MOLECULAR MECHANICS (MM3 AND MM4)
Structures calculated by quantum mechanics differ from those measured experimentally because of the vibrational motions of the molecules. Newer methods have been developed in MM4 that permit a more accurate interconversion of these types of structures.
The enlarged parameter set available in MM4 extends its capabilities to more classes of small molecules encountered in pharmaceutical or agrochemical research programs.
MM4 can carry out calculations on coordination compounds, including square planar for four coordinate; trigonal bipyramid and square pyramid for five coordinate; and octahedral bipyramid for six coordinate compounds. Conjugated p-systems are analyzed by a Variable Electronegativity SelfConsistent Field (VESCF) calculation to determine bond orders and lengths both initially and throughout optimization of the surrounding structure.
MOLECULAR MECHANICS (MM3 AND MM4)
MM4's treatment of hydrogen bonding is improved through the use of special van der Waals distances and a directional character in the hydrogen bonding function that accurately accounts for its anisotropic nature. Other specialized intramolecular interactions, such as the anomeric and trans lone-pair (Bohlmann) effect are reproduced accurately.
MM4 considers not only charge-charge, charge-dipole, and dipole-dipole interactions, but also calculates dipole-induced dipoles. These are important in molecules containing many highly polar groups, such as peptides and polyols. Unlike other force fields, the MM4 charge distribution on each atom depends on the conformation since each conformation has different induced dipoles.
Huckel-MO-calculator
SHMO (Simple Hückel Molecular Orbital Calculator)SHMO is an interactive program to perform electronic structure
calculations within the "Simple Hückel Molecular Orbital" approximations. The theoretical basis for the method is described in "Orbital Interaction Theory of Organic Chemistry", Second Edition by Arvi Rauk (Wiley Interscience, 2001), Chapters 3 and 5.
SHMO is designed as a teaching aid and replaces the earlier version distributed with the book. Besides performing conventional SHMO calculations, SHMO permits easy changes of orbital electronegativities (the Hückel coulomb integrals, alpha, hX) and intrinsic orbital interaction values (the Hückel resonance integrals, beta, kXY) to illustrate the effects of interacting orbital energies orbital, energy differences and overlaps on the resultant molecular orbital energies and polarizations.
Huckel-MO-calculator
The first version of this program has been written entirely in the Java language by Richard Cannings. Some bug fixes and new options, and new display of the data and the molecular orbitals has been programmed by Hans-Ulrich Wagner.
Orbitals are displayed as red/green or light grey/grey coloured circles. Empty oribtals can be displayed with white centers, half filled with yellow centers. This option can be switched using the button "Occup".
Orbital energies are displayed as "alpha + x beta" (the coulomb integral, alpha, and the resonance integral, beta, are both negative).
Orbital selection is by clicking on the orbital energy display or step up/step down buttons.
Huckel-MO-calculator
Heteroatoms, X, are incorporated as aX = alpha + hX beta, and bXY = kXY beta , where the default values of the parameters, hX and kXY, are taken from the publication of F. A. Van-Catledge, J. Org. Chem. 1980, 45, 4801-4802.
Structural formula can be drawn in Hückel which serves as input to the calculation. From the structure formula, Hückel builds the matrices involved and performs the calculations. The resulting energy levels are presented in a level diagram, and a molecular orbital diagram shows the resulting molecular orbitals.
Huckel-MO-calculator
Advanced users will appreciate that Hückel provides access to all the results of the calculations involved. They may be readily output to the printer and/or copied to the Windows clipboard if desired. The number of atoms used in the calculations is virtually unlimited. The speed of the calculation will depend on the size of the molecule and on speed of your PC. With a 133 Mhz Pentium it takes about 45 seconds to calculate a molecule with 75 atoms. Hückel permits task switching during the calculations.
Huckel-MO-calculator
The atoms available in the structure formula editor are defined in a database. This database may be easily modified for custom applications. By allowing the definition of new atoms, the parameters involved in the Hückel calculations can be changed.
MOPAC
The name MOPAC should be understood to mean "Molecular Orbital PACkage".
The origin of the name is unique, and might be of general interest: The original program was written in Austin, Texas. One of the roads in Austin is unusual in that the Missouri-Pacific railway runs down the middle of the road. Since this railway was called the MO-PAC, when names for the program were being considered, MOPAC was an obvious contender.
MOPAC
MOPAC is a general-purpose semiempirical molecular orbital package for the study of solid state and molecular structures and reactions.
The semiempirical Hamiltonians MNDO , AM1, PM3, PM6,RM1, MNDO-d , and PM7 are used in the electronic part of the calculation to obtain molecular orbitals, the heat of formation and its derivative with respect to molecular geometry.
Using these results MOPAC calculates the vibrational spectra, thermodynamic quantities, isotopic substitution effects and force constants for molecules, radicals, ions, and polymers.
For studying chemical reactions, a transition state location routine and two transition state optimizing routines are available.
MOPAC
While MOPAC calls upon many concepts in quantum theory and thermodynamics and uses some fairly advanced mathematics, the user need not be familiar with these specialized topics.
MOPAC is written with the non-theoretician in mind. The input data are kept as simple as possible, so users can give their attention to the chemistry involved and not concern themselves with quantum and thermodynamic exotica.
The simplest description of how MOPAC works is that the user creates a data-file which describes a molecular system and specifies what kind of calculations and output are desired.
The user then commands MOPAC to carry out the calculation using that data-file. Finally, the user extracts the desired output on the system from the output files created by MOPAC.
GAUSSIAN
Gaussian 09 is the latest version of the Gaussian® series of electronic structure programs, used by chemists, chemical engineers, biochemists, physicists and other scientists worldwide.
Starting from the fundamental laws of quantum mechanics, Gaussian 09 predicts the energies, molecular structures, vibrational frequencies and molecular properties of molecules and reactions in a wide variety of chemical environments.
Gaussian 09’s models can be applied to both stable species and compounds which are difficult or impossible to observe experimentally (e.g., short-lived intermediates and transition structures).
GAUSSIAN
A) Comprehensive Investigations of Molecules and Reactions
With Gaussian 09, you can thoroughly investigate the chemical problems that interest you. For example, not only can you minimize molecular structures rapidly and reliably, you can also predict the structures of transition states, and verify that the located stationary points are in fact minima and transition states.
You can go on to compute the reaction path by following the intrinsic reaction coordinate (IRC) and determine which reactants and products are connected by a given transition structure. Once you have a complete picture of the potential energy surface, reaction energies and barriers can be accurately predicted.
GAUSSIAN
B) Predicting and Interpreting Spectra
Spectroscopy is a fundamental tool for investigating molecular structures and properties. However, observed spectra are often difficult to interpret. The results of electronic structure calculations can be vital to this process. For example, predicted spectra can be examined in order to determine peak assignments in observed spectra as well as comparing peak locations and intensities with experimental data. Gaussian 09 can also compute relevant spectroscopic constants and related molecular properties with excellent accuracy.
Gaussian 09 can predict a variety of spectra including IR and Raman, NMR, UV/Visible, Vibrational circular dichroism (VCD), Raman optical activity (ROA), Electronic circular dichroism (ECD), Optical rotary dispersion (ORD), Hyperfine spectra (microwave spectroscopy), Franck-Condon, Herzberg-Teller and Franck-Condon/Herzberg-Teller analyses.
SciQSAR
SciQSAR 2D provides powerful, 2D molecular descriptors for QSAR/QSPR analysis on compound databases SciQSAR 2D is packed with integrated features to make over 600 of these powerful descriptors easily accessible, no matter whether you're working with one molecule or thousands!
It interfaces alchemy 2000 and Chem3D.
Molecular Descriptors
Molecular connectivity
Molecular shape
Topological
Electrotopological (Atom E & H E-States)
Electrotopological bond types
SciQSAR
Reads 2D/3D Structure Input
Smiles and SLN strings
ISIS SKC
Alchemy AL2, Hin, MDL Mol, Mol2, SDF
Use built-in 2D sketcher
SciQSAR
Data Collection
Menu-driven selection of descriptors
Automated processing
Ability to use only pharmacophoric atoms
Ability to renumber atoms "on-the-fly"
See descriptor values at-a-glance
Output selected descriptors to Microsoft Excel
PDB
The Protein Data Bank (PDB) is a repository for the three-dimensional structural data of large biological molecules, such as proteins and nucleic acids.
The data, typically obtained by X-ray crystallography or NMR spectroscopy and submitted by biologists and biochemists from around the world, are freely accessible on the Internet via the websites of its member organisations (PDBe, PDBj, and RCSHB).
The PDB is overseen by an organization called theWorldwide Protein Data Bank, wwPDB.
PDB
The PDB is a key resource in areas of structural biology, such as structural genomics.
Most major scientific journals, and some funding agencies, now require scientists to submit their structure data to the PDB. If the contents of the PDB are thought of as primary data, then there are hundreds of derived (i.e., secondary) databases that categorize the data differently. For example, both SCOP and CATH categorize structures according to type of structure and assumed evolutionary relations; GO categorize structures based on genes.[4]
PDB
Most structures are determined by X-ray diffraction, but about 10% of structures are now determined by protein NMR.
When using X-ray diffraction, approximations of the coordinates of the atoms of the protein are obtained, whereas estimations of the distances between pairs of atoms of the protein are found through NMR experiments.
Therefore, the final conformation of the protein is obtained, in the latter case, by solving a distance geometry problem. A few proteins are determined by cryo-electron microscopy.
SYBYL
It is a tool developed by Certara(Tripos) for molecular modeling and simulation.
SYBYL-X’s science offers unique, competitive advantages in a number of key areas vital for today’s successful discovery research, which are summarized below:
• 3D QSAR: use the power of industry leading CoMFA in a new way to generate novel ideas for R-groups — predict the level of biological activity or potency based on structure-activity data, not just yes/no activity predictions
• Ligand-based virtual screening: search millions of compounds overnight — don’t miss hits because you only screened subsetted portions of your database
SYBYL
Cheminformatics: produce highly focused queries that avoid false positives using rich set of 3D queries; on-the fly conformational searching means you only store a single conformation of your molecules, keeping database size small and very transportable
• Docking: custom tailor and fine-tune docking to a particular receptor site using information like SAR or known poses to improve rank ordering of ligands
SYBYL
SYBYL‐X Suite includes the following : Core technologies 1. SYBYL‐X Base 2. Concord Sketch 3. MOLCAD
Advanced Computation Pharmacophore Perception 4. GALAHAD 5. Tuplets 6. GASP 7. DISCOtech8. Distill
QSAR Technologies 10. QSAR with CoMFA 11. Advanced CoMFA 12. HQSAR 13. cLogP/CMR
SYBYL
Protein Modeling & Structure Prediction 14. Biopolymer 15. Advanced Protein Modeling16. Virtual Screening/Docking 17. Surflex‐Dock 18. Cscore
Virtual Screening/ R‐groups 19. Topomer CoMFA 20. Topomer Search Combinatorial Chemistry 21CombiLibMaker/Legion* 22. DiverseSolutions* 23. Selector
SYBYL
ChemInformatics 23. UNITY Base & 3D 24. Concord Standalone** 25. StereoPlex* Value Added Tools 26. Confort 27. ProtoPlex** 28. Surflex‐Si
SYBYL
SAR and QSAR
Certara (then Tripos) was an early pioneer in QSAR (in particular 3D QSAR), and continues to build on that tradition of scientific innovation with key QSAR techniques explicitly designed to advance Lead Optimization:
Comparative Molecular Field Analysis (CoMFA) was the first 3D QSAR method, and is an industry standard with literally thousands of literature publications demonstrating CoMFA’s utility for molecular discovery. With CoMFA, researchers build statistical and graphical models that relate the chemical and biological properties of molecules to their 3D structures and the 3D steric and electrostatic properties. These models are then used to predict the properties or activity of novel compounds.
Hologram QSAR (HQSAR) is a novel 2D QSAR method, and uses counts of key molecular substructures and PLS to generate fragment-based structure-activity relationships. Validation studies have shown that HQSAR has predictive capabilities comparable to those of much more complicated 3D-QSAR techniques.
SYBYL
With Topomer CoMFA, the next generation of CoMFA/3D QSAR, Certara is revolutionizing 3D QSAR all over again. Building on the innovation originated in the traditional CoMFA method, Topomer CoMFA minimizes the preparation needed for 3D QSAR analysis. These models can be created in minutes, and can easily be used by both QSAR experts and QSAR non-experts. Because pose generation is automated with Topomer CoMFA, researchers can:
Easily generate models for multiple biological endpointsIdentify novel ideas for R-groups that are most likely to lead to
improvements in activity using virtual screening based on predictions from 3D QSAR models
Generate 100’s to 1000’s of predictive QSAR models for chemogenomic studies automatically by mining large databases of chemical and biological data
SYBYL
Ligand Based Virtual Screening and Shape Based Similarity and Alignment, Pharmacophore Searching
The SYBYL-X Suite includes a comprehensive set of tools for shape and pharmacophore based analysis and virtual screening:
Topomer Search is an exceptionally fast 3D ligand-based virtual screening tool that has been demonstrated to be effective for both lead hopping and scaffold hopping.
Topomer Search can search millions of structures overnight on a single processor, allowing you to screen very large collections of compounds and avoid the risk of missing important leads because of subsetting. Screen for whole molecules, side chains, or scaffolds using conformationally independent topomer similarity.
SYBYL
Surflex-Sim performs rigorous, flexible 3D shape based virtual screening. In addition, the molecular alignments and hypotheses of bioactive ligand conformations generated by Surflex-Sim stimulate 3-D ligand-based design. Surflex-Sim's comprehensive shape comparison algorithms include the consideration of the molecules’ shape, H-bonding, and electrostatic properties. Shape similarity, as computed by Surflex-Sim has been shown to be highly effective as means of rationalizing and predicting both on- and off-target pharmacological effects.
UNITY-3D is an industry standard with literally thousands of literature publications demonstrating successful lead discovery based on UNITY's conformationally flexible pharmacophore based searching. UNITY's rich set of pharmacophore features enable highly specific queries to be constructed from a lead structure, a pharmacophore model, or a receptor active site.
SYBYL
Galahad, GASP, and DiscoTech are pharmacophore hypothesis generation tools that deduce the spatial requirements for drug binding when a structure of the drug’s target isn’t known. The pharmacophore models that are generated are useful for virtual screening and for inspiring and testing new ideas to see how they match to a set of lead drug candidates
SYBYL
Structure Based Design
The SYBYL-X Suite includes : Surflex-Dock™ offers unparalleled enrichments in virtual
high-throughput screening and accurate prediction for ligand binding mode and conformation.
Surflex-Dock has been extensively validated and favorably compared to all leading competing methods by independent researchers along with numerous studies by the application's author, Prof. Ajay N. Jain, Ph.D., a faculty member at the University of California San Francisco Cancer Research Institute. Its unique capabilities allow users to customize scoring functions, consider protein flexibility, multiple binding site detection and docking, and key fragment constraints and make Surflex-Dock a top choice for lead discovery, scaffold replacement, lead optimization and fragment based discovery.
SYBYL
Advanced Protein Modeling (APM) for protein homology modeling enables you to perform both homolog finding and comparative modeling through a streamlined interface. APM finds more relevant homologs to start the modeling process because in takes advantage of environment-specific substitution tables, structure-dependent gap penalties, automated alignment method selection, and the highly annotated HOMSTRAD database.
Homologs are structurally aligned based using homology and local structural environment, and then structurally conserved regions are identified using backbone curvature and torsion in addition to C-alpha rmsd and homology.
Loops (structurally variable regions) are then modeled by knowledge-based or ab initio approaches, and sidechains are added by enumerating rotamer combinations constrained by borrowing as much information from the parent homologs as possible. A range of analysis tools are available to highlight potential problems with the structure and allow the user to iterate through the process and refine initial models.
SYBYL
MOLCAD exploits the power of the human eye by
creating graphical images that reveal the properties of molecules essential for molecular recognition. MOLCAD calculates and displays the surfaces of both small molecule drugs and macromolecular targets and can identify channels and cavities that may be potential binding sites. A broad range of properties can be mapped onto these surfaces to rationalize the properties of the binding site and ligand/receptor complementarity.
CHARMM force field
CHARMM (Chemistry at HARvard Macromolecular Mechanics) is the name of a widely used set of force fields for molecular dynamics as well as the name for the molecular dynamics simulation and analysis package associated with them.
The CHARMM force fields for proteins include: united-atom (sometimes called "extended atom")
CHARMM19,all-atom CHARMM22 and its dihedral potential corrected variant CHARMM22/CMAP.
In the CHARMM22 protein force field, the atomic partial charges were derived from quantum chemical calculations of the interactions between model compounds and water. Furthermore, CHARMM22 is parametrized for the TIP3P explicit water model.
Nevertheless, it is frequently used with implicit solvents. In 2006, a special version of CHARMM22/CMAP was reparametrized for consistent use with implicit solvent GBSW.
CHARMM force field
For DNA, RNA, and lipids, CHARMM27 is used. Some force fields may be combined, for example CHARMM22 and CHARMM27 for the simulation of protein-DNA binding. Additionally, parameters for NAD+, sugars, fluorinated compounds, etc. may be downloaded.
These force field version numbers refer to the CHARMM version where they first appeared, but may of course be used with subsequent versions of the CHARMM executable program. Likewise, these force fields may be used within other molecular dynamics programs that support them.
CHARMM force field
In 2009, a general force field for drug-like molecules (CGenFF) was introduced. It "covers a wide range of chemical groups present in biomolecules and drug-like molecules, including a large number of heterocyclic scaffolds".
The general force field is designed to cover any combination of chemical groups. This inevitably comes with a decrease in accuracy for representing any particular subclass of molecules.
Users are repeatedly warned in Mackerell's website not to use the CGenFF parameters for molecules for which specialized force fields already exist (as mentioned above for proteins, nucleic acids, etc.).
HYPERCHEM
HyperChem is a sophisticated molecular modeling environment that is known for its quality, flexibility, and ease of use.
Uniting 3D visualization and animation with quantum chemical calculations, molecular mechanics and dynamics, HyperChem puts more molecular modeling tools at your fingertips than any other Windows program.
It includes all the components of structure, thermodynamics, spectra, and kinetics.
HYPERCHEM
DisplayRendering choices: Ball-and-stick, fused CPK
spheres, ball and cylinders, or tubes with optional shading and highlighting. Also vdW dots added to any rendering.
Ribbon rendering for protein backbones, with optional sidechain display.
Cylinders, ribbon lines, thin solid ribbons, thick ribbons and coils for secondary structure rendering.
3D Isosurfaces or 2D contour plots of: Total charge density.
Molecular orbitals, Spin density, Electrostatic potential (ESP).
HYPERCHEM
ESP mapped onto 3D charge density surface Isosurface rendering choices: wire mesh,
JorgensenSalem, transparent and solid surfaces, Gouraud shaded surface. User-specified grid and isosurface value.
Generate ray-traced graphical images. During simulations, display and average kinetic,
potential, and total energy, as well as values of user-specified bond lengths, bond angles, or torsion angles.
Spectra display of IR or UV-VIS. Animate vibrational modes.
NMR spectra. Crystal structures. Slides (molecules plus annotations).
HYPERCHEM
Predict Relative stabilities of isomers Heats of formation Activation energies Atomic charges HOMO-LUMO energy gap Ionization potentials Electron affinities Dipole moments Electronic energy levels MP2 electron correlation energy CI excited state energy
HYPERCHEM
Transition state structures and properties Non-bonded interaction energy UV-VIS absorption spectra IR absorption spectra Rate constants - unimolecular or bimolecular reactions Equilibrium as a function of temperaure Isotope effects on vibrations Collision effects on structural properties Stability of clusters Shielding and coupling constants Conformations of flexible systems Homologous proteins
HyperChem
Save Results Use Import/Export option to save results of
quantum mechanics calculations or to view results generated by other programs.
Use HyperChem Data to store structures and properties in a custom molecular database.
Create Reaction Movies in AVI format
MM+
Merck Molecular Force Field (MMFF) is a family of force fields developed by Merck Research Laboratories.
They are based on the MM3 force field. MMFF is not optimized for a single use (like either simulating proteins or small molecules), but tries to perform well for a wide range of organic chemistry calculations. The parameters in the force field have been derived from computational data.
The first published force field in the family is MMFF94.
MM+
Screening of tens of thousands of compounds based on pharamcophoric models or “interactions” with the binding site of a target molecule. Accordingly, parameters for a wide variety of chemicals are required.
Class I force fields have limited transferability, therefore are only parameterized for a limited number of chemical types and are of limited value for screening large numbers of compounds.
Class II force fields, with their greater transferability, allow for a large number of compounds to be treated, as required for database screening.
MM+
MMFF: A Class II force field designed to be a transferable force field for pharmaceutical compounds that accurately treats conformational energetics and nonbonded interactions. This would, ideally, produce a force field that was adequate for both gas phase and condensed phase calculations.
Transferability: Application of empirical force field parameters to molecules not explicitly included during the parameter optimization.
AMBER
AMBER (an acronym for Assisted Model Building with Energy Refinement) is a family of force fields for molecular dynamics of biomolecules originally developed by Peter Kollman's group at the University of California, San Francisco.
AMBER is also the name for the molecular dynamics software package that simulates these force fields.
The term "AMBER force field" generally refers to the functional form used by the family of AMBER force fields. This form includes a number of parameters; each member of the family of AMBER force fields provides values for these parameters and has its own name.
AMBER Force Field Equation
AMBER
The meanings of right hand side terms are: First term (summing over bonds): represents the energy between
covalently bonded atoms. This harmonic (ideal spring) force is a good approximation near the equilibrium bond length, but becomes increasingly poor as atoms separate.
Second term (summing over angles): represents the energy due to the geometry of electron orbitals involved in covalent bonding.
Third term (summing over torsions): represents the energy for twisting a bond due to bond order (e.g. double bonds) and neighboring bonds or lone pairs of electrons. Note that a single bond may have more than one of these terms, such that the total torsional energy is expressed as a Fourier series.
Fourth term (double summation over and ): represents the non-bonded energy between all atom pairs, which can be decomposed into van der Waals (first term of summation) and electrostatic (second term of summation) energies.
AMBER
To use the AMBER force field, it is necessary to have values for the parameters of the force field (e.g. force constants, equilibrium bond lengths and angles, charges). A fairly large number of these parameter sets exist, and are described in detail in the AMBER software user manual. Each parameter set has a name, and provides parameters for certain types of molecules.
Peptide, protein and nucleic acid parameters are provided by parameter sets with names beginning with "ff" and containing a two digit year number, for instance "ff99".
GAFF (General AMBER force field) provides parameters for small organic molecules to facilitate simulations of drugs and small molecule ligands in conjunction with biomolecules.
The GLYCAM force fields have been developed by Rob Woods for simulating carbohydrates.
AMBER
The AMBER software suite provides a set of programs for applying the AMBER forcefields to simulations of biomolecules. It is written in Fortran 90 and C with support for most major Unix-like systems and compilers. Development is conducted by a loose association of mostly academic labs. New versions are generally released in the spring of even numbered years; AMBER 10 was released in April 2008. The software is available under a site-license agreement, which includes full source, currently priced at US$400 for non-commercial and US$20,000 for commercial organizations.
Programs LEaP is used for preparing input files for the simulation programs Antechamber automates the process of parameterizing small organic molecules
using GAFF SANDER (Simulated Annealing with NMR-Derived Energy Restraints) is the
central simulation program and provides facilities for energy minimization and molecular dynamics with a wide variety of options
pmemd is a somewhat more feature-limited reimplementation of sander by Bob Duke. It was designed with parallel processing in mind and has significantly better performance than sander when running on more than 8–16 processors
AMBER
pmemd.cuda has been made in order to run simulations on GPU enabled machines
pmemd.amoeba was developed to handle the extra parameters in the polarizable AMOEBA force field.
nmode calculates normal modes ptraj provides facilities for numerical analysis of simulation results.
AMBER does not include visualization capabilities; visualization is commonly performed with VMD. Ptraj is now unsupported as of AmberTools 13.
cpptraj is a rewritten version of ptraj made in C++ to provide faster analysis of simulation results. Several actions have been made parallelizable with OpenMP and MPI
MM-PBSA allows for implicit solvent calculations on snap shots from molecular dynamics simulations
NAB is a built in nucleic acid building environment made to aid in the process of manipulating proteins and nucleic acids where an atomic level of description will help with computation.
MOE
MOE, the Molecular Operating Environment, is a comprehensive software system for Life and Material Science developed by Chemical Computing Group Inc. (CCG).
MOE is a combined Applications Environment and Methodology Development Platform that integrates visualization, simulation and application development in one package.
MOE strongly supports drug design through molecular simulation, protein structure analysis, data processing of small molecules, docking study of proteins and small molecules, and so on under the unified operations.
SVL, the Scientific Vector Language, is the portable high performance programming language built-in MOE. SVL is the vectrized command language, scripting language and applications programming language.
One-tenth reductions in code size over C and FORTRAN are routinely realized. SVL source code to MOE applications is distributed with MOE to end-users; this allows MOE applications to be freely customized or modified.
MOE
MOE
Structure-Based Design
Active Site Detection
Protein:Ligand Interaction Diagrams
Streamlined Interface for Ligand Optimization
Contact Statistics, Electrostatic, and Interaction Maps
Solvent Analysis with 3D-RISM
Scaffold Replacement
Ligand:Receptor Docking
Multi-Fragment Search
Automated Structure Preparation
MOE
Fragment-Based Design
Scaffold Replacement
Ligand Hybridization
Fragment Linking and Growing
Medicinal Chemistry Transformations
Automatic Filtering and Scoring
Fragment Databases
Pharmacophore Discovery
Ligand- and Structure-Based Query Editor
Automatic Pharmacophore Generation
Conformation Databases
Virtual Screening
Ligand- and Structure-Based Scaffold Replacement
MOE
Medicinal Chemistry Applications
Streamlined Interface for Structure-Based Ligand Optimization
Visualize and Analyze Non-Bonded Interactions
Protein-Ligand Interaction Diagrams
Surfaces and Maps
Conformational Search and Analysis
Flexible Alignment of Multiple Molecules
Scaffold Replacement, Growing and Fragment Linking
Pharmacophore Discovery
SAReport
Molecular Descriptors for QSAR
Virtual Library Builder
MOE
Biologics Applications
Whole Protein and Interface Visualization and Analysis
2D Protein Interaction Diagrams
Mutation and Rotamer Exploration
Sequence Analyzer and Editor
Protein Engineering
Antibody and Fusion Protein Modeler
Protein Properties
Advanced Molecular Simulations
Molecular Modeling and Simulations
Automated Structure Preparation
Molecular Mechanics
Molecular Dynamics
MOE
Conformational Search and Analysis
Quantum Mechanical and Semi-Empirical Calculations
Flexible Alignment of Multiple Molecules
Protein and Antibody Modeling
Protein Structure Databases
Remote Homology and Fold Identification
Multiple Sequence/Structure Alignment and Analysis
Mutation and Rotamer Exploration
Protein and Antibody Structure Prediction
Advanced Loop Conformation Generation
Protein Geometry Quality Assessment
Automated Structure Preparation
MOE
Cheminformatics and QSAR
Powerful Database Viewer and Workflow Integration
Import, Prepare, Filter, Sort, Merge and Export Molecular Data
Combinatorial Library Design
Molecular Descriptors
QSAR and QSPR Modeling
Similarity, Diversity, and Fingerprints
SAReport
Methods Development and Deployment
Platform Independent
Scientific Vector Language
DISCOVERY STUDIO
Discovery Studio® is a single unified, easy-to-use, graphical interface for powerful drug design and protein modeling research.
Discovery Studio contains both established gold-standard applications (e.g., Catalyst, MODELER, CHARMm, etc.) with years of proven published results, as well as and cutting-edge science to address today’s drug discovery challenges.
Discovery Studio is built on the SciTegic Pipeline Pilot Scitegic Enterprise Server platform™ open operating platform, allowing seamless integration of protein modeling, pharmacophore analysis, and structure based design, as well as third-party applications.
PLATFORMS
Discovery Studio Standalone Discovery Studio Standalone is a complete molecular
modeling platform designed for the independent modeler. This standalone environment, powered by the Pipeline
Pilot open platform, includes the entire infrastructure needed to design and run modeling experiments with Discovery Studio Science.
Easily visualize, model, and analyze biological and chemical data using tools for sketching 3D molecules, visualizing dynamic changes, 3D graphing and a host of other functionality.
The standalone installation can be connected to a Pipeline Pilot Server for easy sharing of workfl ows and data. This confi guration also includes access to a rich set of perl-based scripting commands
PLATFORMS
Discovery Studio Visualizer Client Discovery Studio Visualizer Client is a powerful graphical
interface to access Discovery Studio Science. DS Visualizer Clients can be inexpensively deployed and easily
maintained for research teams while still offering world-class science.
When deployed with a Pipeline Pilot Server, DS Visualizer Client offers unparalleled capabilities for sharing data, workfl ows, and computational resources.
A rich set of perl-based scripting commands, and customized scripts, are also available for the automation and customization of common modeling operations that replicate UI actions or act on DS data models.
Some examples of the large collection of scriptable actions include molecular overlay, chirality and valency checks, prediction of DSC secondary structure, access to electrostatics tools, management of constraints/restraints, and display of surfaces
PLATFORMS
Pipeline Pilot Server Pipeline Pilot makes the most of your information
through industrial-scale data flow control and powerful mining capabilities.
You can graphically compose data processing networks, using hundreds of different configurable components for operations such as data retrieval, manipulation, computational filtering, and display.
These protocols are automatically captured as you create them and you can even publish them for enterprise level deployment.
Your colleagues have the option to invoke your protocols and run them using their own data using a simple Web interface.
PLATFORMS
DS Developer Client This scaled-down version of the Pipeline Pilot client allows you to
customize pre-existing Discovery Studio work flows, and also includes the necessary component collections to enable customization.
Add or remove available components from existing DS protocols/ components. Create a new component and add to an existing DS protocol. Increase automation of DS Protocols (e.g. automate DS protocol to run over all files in a specific folder).
Easily connect to an external relational database. Integrate third-party algorithms (e.g., CORINA, proprietary codes for descriptor calculations, etc.). The customized protocols can be run directly from the DS Developer Client, or saved into a user area and run from the DS interface for interactive modeling.
A free and extensive library of customized components and protocols for Discovery Studio is available at no additional cost at the Accelrys Forums.
PLATFORMS
Discovery Studio Free Visualizer This free, easy-to-use visualization tool is an ideal
solution for managers and researchers who need to collaborate with modelers, but do not need access to the expert-level analysis tools in Discovery Studio.
Many common tasks and commands are available within the Perlbased scripting API, enabling automation and customization of common modeling tasks.
View and share protein and small-molecule data in a clear and consistent way, and in a wide variety of industry-standard formats.
PLATFORMS
ActiveX Control The Discovery Studio Visualizer ActiveX Control is a
fully integrated, 3D molecular renderer for the Windows environment, and provides a powerful method of sharing scientific results with colleagues via presentations, web pages, etc.
The Control can be inserted into any application that can host a COM component, such as Internet Explorer, PowerPoint, or VC clients.
The OpenGL capabilities of ActiveX Control allow you to render sophisticated molecular images.
In addition to the Viewer’s native format, the ActiveX Control can read many industry standard molecular fi le formats.
Applications
Protein Modeling and Sequence AnalysisBiopolymer Building and AnalysisSimulationsStructure-Based DesignLigand-Based DesignPharmacophore Modeling and AnalysisQSAR and Library DesignADMET
FlexX
FlexX is a computer program for predicting protein-ligand interactions.
For a given protein and a ligand, FlexX predicts the geometry of the complex as well as an estimate for the strength of binding. In this first version of FlexX, the protein is assumed to be rigid.
Thus, the protein must be given in a conformation which is similar to the bound state. The docking algorithm in FlexX works without manual intervention. Nevertheless, in some cases additional information about the ligand or even the complex is known.
You can integrate this knowledge in the computations with FlexX by performing single steps manually.
Thus, FlexX is ideal for interactive work on protein-ligand complexes as well as for screening a larger set of ligands in order to find new leads for drug design.
In summary, FlexX can be useful in the following situation: You have a good three-dimensional model of the protein and you know the location of the active site. You have a set of ligands and you want to know whether and how each of them binds to your protein model.
FlexX
Major workflows with FlexX is: 1. Preparation of the binding site using the Receptor Intelligence of
the Receptor Preparation Wizard This includes selection of chains, receptor protonation, tautomers, etc.
2. Optional: Definition of a receptor-based pharmacophore 3. Read in of a ligand “Docking Library” (one or more ligands) 4. Docking including possibly altering of parameters and further
cycles of docking The character of this workflow is driven by few ligands and/or a
preparational stage. It can easily and comfortably be accomplished with LeadIT .
A Receptor, once prepared with the Wizard, can be saved and read into the commandline mode of FlexX.
Precondition: You must have one or more ligands in a low-energy conformation ready to dock. FlexX does not minimize ligands before docking. The docking process deals with the translational, torsional, and ring conformation degrees of freedom only.
FlexX
Scientifically, FlexX is currently further developed at three locations, at BioSolveIT GmbH (Sankt Augustin), at the ZBH of the University of Hamburg (Hamburg) and at the Max-Planck-Institute for Computer Science (Saarbrücken).
Current further developments of the software:FlexX-PVM: Automatic scheduling on large compute clustersFlexX-C: Efficient docking of combinatorial librariesFlexX-PHARM: Docking under pharmacophore peripheral
conditionsFlexX-Ensemble (FlexE): Docking in consideration to protein
flexibilityFlexX-Scan: Fast Structure-Based Virtual Screening using a
novel receptor binding site descriptorFlexX Metal Interaction: Automatic calculation and definition of
metal coordination geometries
Introduction to CCDC GOLD Suite
• CCDC- Cambridge Crystallographic Data Centre• GOLD (Genetic Optimisation
for Ligand Docking) is a genetic algorithm for docking flexible ligands into protein binding sites.
• Predicting how a small molecule will bind to a protein is difficult, and no program can guarantee success. The next best thing is to measure as accurately as possible the reliability of the program, i.e. the chance that it will make a successful prediction in a given instance.
GOLD
• GOLD is supplied as part of the GOLD Suite, which includes two additional software component, Hermes and GoldMine.
The Hermes visualiser can be used to assist the preparation of input files for docking with GOLD, visualisation of docking results and calculation of descriptors. The Hermes visualiser is also used for interactive docking setup, e.g. for defining the binding site and the setting of constraints.
GoldMine is a tool for the analysis and post-processing of docking results
Using the GOLD Wizard to Prepare the Protein File
Selecting a Protein In the Select proteins to use window, read in
the protein file, 3S7S.pdb, by hitting the Load Protein button,
Select the file Click on Open.
STEPS
Add hydrogens: All hydrogen atoms must be present in the protein input file The hydrogen atoms are placed on the protein in order to ensure that ionisation and tautomeric states are defined unambiguously. Advanced options within GOLD allow for switching between different tautomeric states during docking.
Delete Waters: Water molecules often play key roles in protein-ligand recognition. Water molecules can either form mediating hydrogen bonds between protein and ligand, or be displaced by the ligand on binding. Water molecules within the active site can be retained and allowed to toggle (i.e. switch on and off during docking) and rotate to optimise their H-bonding positions. Those outside of the binding site can be removed from the protein altogether.
Delete ligands The 3S7S.pdb protein is the raw PDB file which is the original protein-ligand complex. For GOLD to effectively dock a ligand back into the active site, the co-crystallised ligand must first be removed.
Defining the Protein Binding Site
It is necessary to specify the approximate centre and extent of the protein binding site, this can be done in a number of ways from within the Define the binding site window, including from :
A protein atomA point A reference ligand A file containing a list of atoms or residues
Specifying a Configuration File Template
At this point you are given the option to load a configuration file template. Configuration templates can be used to load recommended settings for a number of different types of docking protocols
Click Next to proceed to the Select ligands step.
Specifying the Ligand File
From within the Select Ligands window it is possible to Add single ligands
Select a complete directory of ligand files.Specify a single file containing several ligands
(i.e. a multi-MOL2 or SD file).Click Next to proceed to the Choose a fitness
function window.
Selecting a Fitness Function
GOLD offers a choice of scoring functionsGoldScoreChemScoreAstex Statistical PotentialCHEMPLPUser Defined Score
Selecting Docking Speed
GOLD optimises the fitness score using a genetic algorithm (GA).
A number of parameters control the precise operation of the genetic algorithm. The settings are encapsulated into three speeds:
Slow (most accurate): this equates to 100,000 operations
Medium: 50,000 operationsFast (least accurate): 10,000 operations
STEPS
Finished our docking setup and re presented with a Run GOLD button with which we can start the docking.
If we were to click on Run GOLD now, the output would be written to the directory the 3S7S.pdb file is stored in. It is generally preferable to write output to a separate directory.
LIGANDS
PDB ID: 3S7SLigands taken
OO2N
NH
S O
O
OO2N
N
S O
O
CH3
O
CH3
O2N
N
S OO
CH3
O
CH3
O2N
NH
S OO
OO2N
NH
S OO
OO2N
NS
O
O
CH3
O
O2N
HN S
O
OOO2N
NS
O
O
CH3
Mol 01 Mol 02 Mol 03 Mol 04
Mol 05 Mol 06 Mol 07
Mol 08
SCHRODINGER
Schrödinger software suite is a drug design software using both ligand and structure-based methods. Schrödinger provides accurate, reliable, and high performance computational technology to solve real-world problems in life science research. It provides superior solutions and services for the design, selection, and optimization of novel drug candidates.
Schrödinger's predictive models will enable drug discovery scientists to assess properties of chemical compounds early in the discovery process and to select drug candidates that have optimal profiles.
The predictive power of Schrödinger's software allows scientists to accelerate their research and development activities, reduce research costs, and make novel discoveries that might not be possible with other computational or experimental approaches.
Maestro is a powerful, all-purpose molecular modeling environment. Maestro is the unified interface for all Schrödinger software products.
MODULES
Phase Maestro MarcModelPrimeJaguarGlide LiaisonQsiteImpact
Phase- pharmacophore generation module.Maestro- Schrodinger modelling interface. Macro modeller- basic molecular modelling
package (build/ edit functions, minimization and dynamics).
Prime- protein homology modeller (secondary structure prediction combined with homology modelling).
Jaguar- ab initio modelling software (quantum mechanical calculations for small molecules).
Glide- ligand docking program (drug screening).Liason- ligand binding affinity software.Qsite- combined quantum mechanics/ molecular
modelling package (investigation of enzyme active sites).
Glide
Glide is designed to assist in high-throughput screening of potential ligands based on binding mode and affinity for a given receptor module.
One can compare ligand scores with those of other test ligands or compare ligand geometrics with those of a reference ligand.
Additionally one can use Glide to generate one or more plausible binding modes for a newly designed ligand.
Once favourable structures and/or bonding conformations are located with Glide one may use Liaison to obtain binding energies for ligand-receptor pairs.
liasion
Liaison is designed to predict ligand binding energies.
Ligands for which binding energies are known are used to calculate the Liaison SVD-fitted parameters-alpha, beta, and gamma.
These parameters are subsequently used to predict binding energies for other ligands with the same receptor.
QSite
QSite is mixed mode Quantum Mechanics/Molecular Mechanics program to study geometries and energies of structures not parameterized for use with molecular mechanics, such as those that contain metals or transition states.
Qsite is uniquely equipped to perform QM/MM calculations because it combines the superior speed and power of Jaguar.
Maestro
Maestro is the graphical user interaface for all Schrödinger's products.
FirstDiscover (Glide, Impact, Liaison and Qsite) Jaguar, MacroModel etc.
It contains tools for building, displaying, and manipulating chemical structures, for organizing, loading and storing these structures and associated data and for setting up, monitoring, and visualizing the results of calculations on these structures.
Impact
Impact is the computational engine that can perform molecular mechanics calculations either through the maestro interface or from the command line
Jaguar
Jaguar is designed to increase the speed of ab initio calculations in order to accelerate basic and applied research projects and to enable calculations at a higher level of theory.
Jaguar's speed and power make it possible to study larger systems than ever before, or to study many more systems than previously possible, within a reasonable time frame.
MarcModel
MarcModel, with a large selection of force fields and advanced methods for conformational analysis, molecular dynamics, and free energy calculations, is the most trusted name in molecular mechanics.
MarcModel is especially well suited for general-purpose molecular mechanics for small and medium-sized organic molecules in both gas and solution phases.
Further, MarcModel has powerful utilities for exploring proteins and protein-ligand complexes.
PRIME
Prime is highly accurate protein structure prediction suite of programs that integrates comparative modelling and fold recognition into a single user friendly, wizard like interface.
Prime SP (Structure Prediction) consists of a series of steps leading from a protein sequence through comparative modelling to the construction of a 3D structure.
CHEM3D
Chem3D is a molecular modeling package.It performs calculations using MM2 and extended
Huckel as well as acting graphic interface for MOPAC or GAUSSIAN.
There are also plug-ins available for viewing structures and surfaces.
Integrated with molecular analysis, this modeling, visualization and analysis suite provides simultaneous 2-D and 3-D editing, automatic display of hydrogen bonds in the 3-D view, graphical display distance and angle measurements in the 3-D view that are linked to the measurements table, and the ability to compute electronic properties with MOPAC, GAMESS, Gaussian and Jaguar.
Creating, Opening, and Importing Models
Chem3D provides the following starting points for modeling:
Creating a model using the Chem3D toolsOpening an existing 3D modelCreating a model from stationery or templates Importing files in other chemistry file formats
Creating, Opening, and Importing Models
Chem3D provides the following starting points for modeling:
Creating a model using the Chem3D toolsOpening an existing 3D modelCreating a model from stationery or templates Importing files in other chemistry file formats
MM2 in Chem3D
The Chem3D MM2 menu provides computations using the MM2 force field.
The MM2 procedures described assume that you understand how the potential energy surface relates to conformations of your model.
The energy minimization routine performs a local minimization only. Therefore, the results of minimization may vary depending on the starting conformation in a model.
MOPAC in Chem3D
MOPAC is a molecular computation application developed by Dr. James Stewart and supported by Fujitsu Corporation that features a number of widely-used, semi-empirical methods. It is available in two versions, Standard and Professional.
MOPAC Std. is available as an optional addin for Chem3D Pro. MOPAC Pro is a standard feature of Chem3D Ultra.
Chem3D MOPAC provides a graphical user interface (GUI) that allows you to perform MOPAC computations directly on the model in the Chem3D model window. As a computation progresses, the model changes appearance to reflect the computed result. Computations are limited in CS MOPAC to 250 heavy atoms.
GAMESS in Chem3D
The General Atomic and Molecular Electronic Structure System (GAMESS) is a general ab initio quantum chemistry package maintained by the Gordon research group at Iowa State University. It computes wavefunctions using RHF, ROHF, UHF, GVB, and MCSCF. CI and MP2 energy corrections are available for some of these.
GAMESS is a command-line application, which requires a user to type text-based commands and data. Chem3D serves as a front-end graphical user interface (GUI), allowing you create and run GAMESS jobs from within Chem3D.
Descriptors of SAR
Chem3D provides a set of physical and chemical property predictors. These predictors, which help predict the structure-activity relationship (SAR) of molecules, are referred to as SAR descriptors. The Chem3D Property Broker provides an interface in Chem3D and ChemSAR/Excel that allows you to calculate properties using many calculation methods provided by various Property Server components.
Property broker server in Chem3D
The components of the Property Broker-Server architecture are illustrated below:
ChemProp Std Server
Connolly Solvent Accessible Surface Area Connolly Molecular Surface Area Connolly Solvent-Excluded Volume Exact Mass Molecular FormulaMolecular WeightOvalityFormal Charge
ChemProp Pro Server
Boiling Point Critical TemperatureCritical PressureCritical VolumeMelting PointLogPMolar Refractivity
limitations
Property prediction using CS ChemProp Pro has following limitations:
Single molecules with no more than 100 atoms.Literature values for Partition Coefficients (LogP) and
Henry's Law Constant are not available for all molecules.Some atom arrangements are not parameterized for the
fragmentation methods used to calculate the properties.
MM2 Server
Bending Energy. Charge-Charge EnergyDipole MomentNon-1,4 van der Waals EnergyTorsion EnergyTotal Energyvan der Waals Energy
MOPAC Server
Alpha Coefficients Beta Coefficients Gamma Coefficients Dipole Electronic Energy HOMO Energy LUMO Energy Repulsion Energy Symmetry Total Energy
GAMESS Server
Dipole MomentHOMO EnergyLUMO EnergyRepulsion Energy Energy. Total Energy
SAR Descriptor Computations
Chem3D performs property prediction calculations.
These computed properties are the descriptors that may be used to estimate the structure-activity Relationship(SAR) of molecules.
Selection of properties
To select properties for computation:
1. From the Analyze menu, choose Compute Properties.
The Compute Properties dialog box appears.
2. Set appropriate values for the Class, Server, Cost, and Quality filters.
3. From the list of Available Properties, select the properties to calculate.
4. Click Add.
Property Filters
• Class-limits the list of available properties to types calculations that you specify.
• Server-limits the list of available properties to those properties computed by the servers you specify.
• Cost-represents the maximum acceptable computational cost. It limits the list of available properties to those which are less than or equal to the computational cost specified.
• Quality-represents the minimum acceptable data quality. It limits the list of available properties to those with quality greater than or equal to the quality specified.
Setting Parameters
If a property has one or more parameters that affect the result of the calculation, you can specify the values or calculation method of those parameters. If several properties have the same parameters, you can change the parameters simultaneously.
To change a parameter:
1. Select the property or properties in the Selected Properties list.
2. Click Parameters.
3. Edit the value or select the method.
4. Click OK.
ALCHEMY 2000
It is a graphic interface for running molecular mechanics and semi-empirical calculations.
Calculations can be done with the built-in tripos force field or by calling the MM3 or MOPAC programs, which are included with the package.
Alchemy is designed by Tripos and sold by SciVision.
Molecules can be drawn in 2D or 3D and can also be inter-converted.
Libraries of organic functional groups is available and it also has protein builder.
ALCHEMY 2000
The user can change stereochemistry or conformational angle until he wants it to stop.
Conformational searches using Tripos force field is done up to eight single bonds and two rings.
Output data can be exported to a spreadsheet.Structures can be labeled and rendered in several
different ways and can also be saved in different formats or as an image. The presentation mode allows molecular structure to be combined with text.
The program allows to create a database and run calculations on whole database.
REFERENCE
www.google.comManuals of different softwareswww.sciencedirect.comComputational chemistry by David Young
