COMP Programming

229th ACS National Meeting
San Diego, CA
March 13 - 17, 2005
W. D. Cornell, Program Chair

COMP 1 [809475]: John Pople: The early years

A. David Buckingham, Department of Chemistry, Cambridge University, Lensfield Road, Cambridge CB2 1EW, United Kingdom, Fax: +44 1223 336362, adb1000@cam.ac.uk

I shall comment on John Pople's life in the west of England as a boy, his time as a student at Cambridge University and as a member of the Cambridge Theoretical Chemistry Group, his summer visits to the National Research Council of Canada, and his time as a senior scientist at the National Physical Laboratory in Teddington. I shall also mention my experience as John's first graduate student and as a long-time friend.

COMP 2 [806758]: A half-century of a PP friendship

Robert G. Parr, Department of Chemistry, University of North Carolina, Chapel Hill, NC 27599-3290, rgparr@email.unc.edu

Recollections!

COMP 3 [818384]: John Pople, the early years: From Carnegie Tech to Carnegie-Mellon

Mark S. Gordon, Chemistry Department and Ames Laboratory USDOE, Iowa State University, 201 Spedding Hall, Ames, IA 50011, Fax: 515-294-5204, mark@si.fi.ameslab.gov

Professor Pople's early years, from the beginning at Carnegie Institute of Technology, the development of the CNDO/INDO methods, the original framework of the program that ultimately became Gaussian, and the trnasformation to Carnegie-Mellon University will be discussed.

COMP 4 [827540]: The birth of Gaussian

Warren J. Hehre, Wavefunction, Inc, 18401 Von Karman Avenue, Suite 370, Irvine, CA 92612, Fax: 949-955-2118, hehre@wavefun.com

As was usually the case, John Pople was way ahead of his time. He was worrying about "mflops" years before the phrase came into common use. John's approach to the "two-electron integral problem" in the late 1960's led directly to the birth of Gaussian 70, the first of the "Gaussian" programs, and more importantly to the realization that ab initio (Hartree-Fock) calculations could serve as practical tools for chemists. The underlying idea is simple. Atomic orbitals come in sets ("shells"), for example, a "2p shell" comprising 2px, 2py and 2pz orbitals or a "2sp shell", comprising a 2s orbital in addition to the three 2p orbitals. The individual two-electron integrals resulting from combinations of atomic orbitals on four different shells (256 integrals in the case of four sp shells)share significant information in common, and it is much more efficient to evaluate the full set of integrals all together rather than one integral at a time. John's imaginative algorithm did just this and is what made Gaussian possible.

COMP 5 [803646]: The later years: CMU, Stockholm and beyond

Peter M.W. Gill, Research School of Chemistry, Australian National University, Canberra ACT 0200, Australia, Fax: 61-2-6125-0750, peter.gill@anu.edu.au

I met John in 1984, joined his group in 1988 and remained with him for five years. It was an intensely exciting time for me, as I found myself thrust to the cutting edge of quantum chemical developments, and the scientific momentum that I gained in Pittsburgh has continued to fuel my research activities to the present day. Despite moving from the US, first to New Zealand and later to England, I had the good fortune to retain John and Joy's friendship to the end. It was a privilege to know both the public scientist and the private man.

My lecture will sketch the progress of the last two decades of John's life, outlining the computational highlights and reflecting on the lessons that can be learned from one of the most influential chemists of our time.

COMP 6 [807999]: John in Evanston: Some reminiscences and thoughts

Mark A. Ratner, Department of Chemistry, Northwestern University, 2145 Sheridan Rd., Evanston, IL 60208-3113, Fax: 491 7713, ratner@chem.northwestern.edu

John Pople's scientific career was completed in his years at Northwestern, where he served as Trustee's Professor of Chemistry. During that time, his remarkable scope and insight were made even clearer, as he began work on relativistic quantum chemistry, on image potentials, on metal cluster calculations, on new aspects of DFT calculations and on the entropy of a fried egg, the project that he was pursuing at the time of his death. This was also the period during which his tremendous contributions to the chemical sciences were recognized by the Nobel Foundation in an appropriate (if belated) fashion.The clarity of John's vision, the economy and directness of his science and the warmth and supportiveness of his scientific personality did a great deal to shape the nature of the contemporary chemical sciences. On a smaller scale, they also helped to shape the Northwestern Chemistry Department. This short talk will include both some reminiscences and a comment or two on the science involved!

COMP 7 [845865]: A friendship late in life

Walter Kohn, Department of Physics, University of California, Santa Barbara, Santa Barbara, CA 93106, kohn@physics.ucsb.edu

COMP 8 [831962]: Computers as tools of discovery

Uzi Landman, School of Physics, Georgia Institute of Technology, Atlanta, GA 30332-0430

Dictionary definitions of "to simulate" vary - from "pretend to be" and "imitate or counterfeit", to "produce a computer model of a process". In this talk we will discuss and demonstrate methodological and scientific issues pertaining to the use of computer simulations as faithful representations of complex natural phenomena of predictive power,rather then "imitation and counterfeit". Examples will include: nano-scale fluid dynamics - nanojets; nanocatalysis by supported gold clusters; transport and reactions of ionization holes in DNA; formation and properties of highly correlated electron molecules in two-dimensional quantum dots, and bosonic crystallites in traps.

COMP 9 [805490]: Application of non-Hamiltonian molecular dynamics to chemical systems

Glenn Martyna, Physical Science Division, IBM Research, TJ Watson Research Center, PO Box 218, Yorktown Heights, NY 10598, Fax: 914-945-4506, martyna@us.ibm.com

The theory of non-Hamiltonian dynamical systems and their application to the study of the equilibrium properties of complex chemical processes has been an important topic research in the impressive career of Professor M.L. Klein. Here, the basic non-Hamiltonian molecular dynamics formalism is given followed by the basic equations for stable, ergodic, constant temperature and constant pressure methods. The techniques are then applied within the Car-Parrinello formalism to describe the properties of neat liquid water at two temperatures and the solvation of a peptide fragment, N-methyl acetamide, in water.

COMP 10 [832735]: Stochastic linear scaling for metals and non metals

Michele Parrinello and Florian R Krajewski, Physical Chemistry ETH (Zurich), c/o USI-Campus Via Buffi 13, Lugano 6900, Switzerland, Fax: +41-91-9138-817, parrinello@phys.chem.ethz.ch

Total energy electronic structure calculations, based on density functional theory or on the more empirical tight binding approach, are generally believed to scale as the cube of the number of electrons. By using the localisaton property of the high temperature density matrix we present exact deterministic algorithms that scale linearly in one dimension and quadratically in all others. We also introduce a stochastic algorithm that scales linearly with system size. These results hold for metallic and non metallic systems and are substantiated by numerical calculations on model systems.

COMP 11 [825773]: Probing long time-scale events with advanced simulation techniques

Bin Chen, Department of Chemistry, Louisiana State University, Baton Rouge, LA 70803-1804, Fax: 225-578-3458, binchen@lsu.edu

Despite the recent advent of multi-teraflop computing platforms, long time-scale events such as phase transition and protein folding remain out of reach for conventional simulation methods. The sampling difficulties for these long-time scale events are caused by large free energy barriers and the inherent micro-heterogeneity of the phase space. Although the former problem can now be surmounted by a host of free-energy based methods (including umbrella sampling), a separate approach is still required to deal with the latter problem. For example, for vapor-liquid nucleation the micro-heterogeneity arises from the presence of a spectrum of microphase regions (e.g., monomers and clusters). These microphase regions differ to a great extent on both energetic and entropic factors. In contrast, the random displacements used in the conventional Metropolis Monte Carlo scheme and the force-driven diffusion employed by molecular dynamics lack the balance of these two factors. This leads to slow transfer of particles between the micro-phase regions. A novel technique, called aggregation-volume-bias Monte Carlo (AVBMC), can overcome this problem by explicitly dividing the space surrounding a molecule into the associating and non-associating regions. This allows for direct transfer between microphase regions, thereby bypassing the time and spatial constraints imposed on molecular dynamics and Metropolis Monte Carlo techniques. AVBMC can be combined with configurational-bias Monte Carlo, umbrella sampling, histogram-reweighting, and density of states methods to study nucleation phenomena for complex systems. Applications to nucleation of neat water using polarizable force fields and mixtures containing water, alcohols, and alkanes as well as protein crystallization will be presented.

COMP 12 [833259]: BioSimGrid: A distributed database for the storage and analysis of biomolecular computer simulations

Jonathan W Essex¹, Stuart E Murdock¹, Robert J Gledhill¹, Kaihsu Tai², Muan Hong Ng³, Steve Johnston³, Bing Wu⁴, Hans Fangohr³, Paul Jeffreys⁴, Simon Cox³, and Mark Sansom². (1) School of Chemistry, University of Southampton, Highfield, Southampton SO17 1BJ, United Kingdom, Fax: +44 (0)23 8059 3781, jwe1@soton.ac.uk, (2) Department of Biochemistry, University of Oxford, (3) e-Science Centre, University of Southampton, (4) e-Science Centre, University of Oxford

BioSimGrid aims to deliver a biomolecular simulation data repository to enable more efficient storage, access, exchange and analysis of biomolecular simulation data. The project seeks to exploit the concepts of the Grid, where large computational and data resources are made available to users in a highly accessible manner. The ability to submit, search, query, retrieve, and post-process biomolecular simulation data in a uniform way is important for allowing more efficient data sharing. In this presentation, the design and operation of the current version of the software will be described, and future extensions outlined. The application of BioSimGrid to the comparative analysis of a number of protein simulations (acetylcholinesterase with bacterial outer-membrane phospholipase, and a number of mutants of the prion protein with each other) will also be presented.

COMP 13 [832715]: Binding energy calculation for FKBP receptor and ligands by generalized BAR method

Hideaki Fujitani¹, Yoshiaki Tanida¹, Masakatsu Ito¹, Michael R. Shirts², Christopher D. Snow³, Guha Jayachandran², and Vijay S. Pande². (1) Fujitsu Laboratories Ltd, 10-1 Morinosato Wakamiya, Atsugi, Japan, Fax: 81-46-250-8844, fjtani@labs.fujitsu.com, (2) Department of Chemistry, Stanford University, (3) Biophysics Program, Stanford University

Generalized Bennett acceptance ratio method (BAR; M. Shirts et al. 2003) enable us to calculate the binding energy of realistic complex systems interacting with the explicit waters by the massively parallel computational method. In order to suit the generalized BAR method, we have developed special computer, BioServer, which has 1920 FR-V processors (high performance and low power processor with 8 way VLIW architecture made by FUJITSU) in one rack. Molecular dynamics (MD) calculations can be efficiently performed on BioServer using Gromacs MD package. In order to get accurate binding energy, we used the Amber 1999 force field for the FKBP receptor and new general Amber force field (GAFF; J. Wang et al. 2004) for ligands. We compare our calculated binding energies for nine ligands with experimental inhibition constants, and we show BioServer is suitable to do the computational drug design.

COMP 14 [814623]: Coupled reference interaction site model (RISM)/simulation approach for free energy of solvation

Holly Freedman and Thanh N. Truong, Henry Eyring Center for Theoretical Chemistry/Department of Chemistry, University of Utah, 315 S. 1400 E. Rm 2020, Salt Lake City, UT 84112, holly@mercury.hec.utah.edu

A novel methodology will be discussed for the determination of solvation free energies from molecular simulation. A single molecular dynamics or Monte Carlo simulation is first carried out on an explicitly solvated system of interest. The resulting radial distribution functions of solvent atoms about all solute atoms are used as input to the solvation free energy expression derived from RISM integral equation theory. Thus simulation may be applied to solvation free energy determination in an alternative approach to the accurate but extremely laborious free energy perturbation and thermodynamic integration techniques. Applications to problems including the molecular conformation of the alanine dipeptide, the tautomeric equilibrium of cytosine, and the potential of mean force profile of an SN2 reaction in aqueous solution will be presented.

COMP 15 [831256]: Computational strategy for binding of phenylalanine analogs in phenylalanyl-tRNA synthetase

Peter Michael Kekenes-Huskey¹, Ismet Caglar Tanrikulu¹, Victor Wai Tak Kam¹, Nagarajan Vaidehi², and William A. Goddard III². (1) Department of Chemistry, Caltech, MC 139-74, Caltech Division of Chemistry, Pasadena, CA 91125, huskeypm@wag.caltech.edu, (2) Materials and Process Simulation Center, California Institute of Technology

A computational strategy has been developed that captures the binding characteristics of phenylalanine analogs to the phenylalanyl-tRNA synthetase (pheRS). Binding to this synthetase is a requisite for incorporation into a translated protein, therefore this approach can provide an indirect measure of incorporation for a candidate analog. The protocol couples a ligand torsion sampling algorithm, moleculeGL, with a side chain replacement program, SCREAM, to generate conformations that maximize binding. This protocol has been applied to binding of the twenty natural amino acids and several phenylalanine analogs in select hosts and compared to measured kcat/km data. The reported binding energies correlate well with measured kcat/km data for AMP activation, thus establishing the efficacy of the approach. Furthermore, this strategy has been applied to several amino acid analogs of interest, yielding binding data for wild-type proteins, as well as proposed mutations to improve binding.

COMP 16 [807221]: Computational study of IAG-nucleoside hydrolase: Determination of the preferred ground state conformation and the role of active site residues

Devleena Mazumder Shivakumar, Department of Chemistry & Biochemistry, University of California Santa Barbara, Santa Barbara, CA 93106, devleena@chem.ucsb.edu, and Thomas C. Bruice, Department of Chemistry and Biochemistry, University of California at Santa Barbara

The mechanism of action of inosine–adenosine–guanosine nucleoside hydrolase (IAG-NH) has been investigated by long-term molecular dynamics (MD) simulation in TIP3P water using stochastic boundary conditions and CHARMM forcefield. Special attention has been given to the role of leaving group pocket residues, conformation of the bound substrate at the active site of IAG-NH and the protonation state of the residue Asp40. This is the first time the structure of flexible loop (missing from all the five X-ray structures) in IAG-NH is reported and its role in catalysis highlighted. Five MD studies have been performed with: enzyme substrate complexes: Enzyme•anti-Adenosine with Asp40-COOH [E(40H)•Ade(a)], Enzyme•anti-Adenosine with Asp40-COO^- [E(40)•Ade(a)], Enzyme•syn-Adenosine with Asp40-COOH [E(40H)•Ade(s)], Enzyme•syn-Adenosine with Asp40-COO^- [E(40)•Ade(s)] and Enzyme•anti-Inosine with Asp40-COO^- [E(40)•Ino(a)]. The protonation state of Asp40 obtained from MD simulation is also confirmed by Poisson Boltzmann equation module for continuum electrostatics. Results from all the five MD simulations as well as Normal mode analysis will be discussed in details.

COMP 17 [828482]: Intermolecular potentials of mean force of amino acid side chain interactions in aqueous medium

Sergio A. Hassan, Center for Molecular Modeling, National Institutes of Health, DHHS, Bethesda, MD 20892, mago@helix.nih.gov

Continuum approximations of solvent effects in mesoscopic systems such as proteins and other biopolymers require an accurate description of hydrogen-bonding (HB) interactions in solution. Here, a systematic study of the potentials of mean force (PMF) of all H-bonded amino acid in water is reported. HB partners are classified according to the hybridization states of the donor and acceptor atoms, and the net charge of the interacting pairs. Constrained molecular dynamics simulations are carried out to calculate the intermolecular mean force (MF) in TIP3P water. Long-range forces are calculated using PME summation in a cubic lattice with PBC. Intermolecular PMF are obtained by integrating the MF along a specified reaction path and statistical errors estimated. Results from long (30-100 ns) dynamics simulations of small (50-100 aa) proteins with the SCP-based continuum model of solvent effects are reported. The implications of these results for quantifying protein-protein/ligand interactions and association/dissociation rates are discussed.

COMP 18 [819693]: Prediction of protein geometry and stability changes for arbitrary single point mutations

Andrew Bordner and Ruben Abagyan, Department of Molecular Biology, The Scripps Research Institute, 10550 N Torrey Pines Road, TPC-28, La Jolla, CA 92037

We have developed a method to predict the changes in both the side chain conformations and the stability of proteins due to single point mutations. A procedure using Monte Carlo based molecular mechanics calculations to predict the geometry of the mutated protein was validated on a large set of X-ray structures for pairs of proteins differing by a single point mutation. An empirical energy function, that included contributions from both the folded and denatured protein conformations, was then fit to experimental stability data using the predicted geometry. This set contained a substantial amount of small to large residue mutations not considered by previous studies. The prediction method gave a standard error of 1.1 kcal/mol for 97% of an independent test set. A fit to the stability data using exclusively conformation independent residue parameters suggests mutations for improving protein stability in the absence of structural information.

COMP 19 [833180]: Side-chain conformational changes on ligand binding: An analysis of the PDB

Francesca Toschi¹, Andrew R. Leach², Paul Bamborough², and Jonathan W Essex¹. (1) School of Chemistry, University of Southampton, Highfield, Southampton SO17 1BJ, United Kingdom, Fax: +44 (0)23 8059 3781, jwe1@soton.ac.uk, (2) Computational Chemistry and Informatics, GlaxoSmithKline Research and Development

Ligand binding may involve a wide range of structural changes in the receptor protein, from hinge movements of entire domains to small side-chain rearrangements in the binding pocket residues. In this presentation, the analysis of side-chain dihedral angles for a range of apo-apo, apo-holo, and holo-holo protein pairs is reported. A range of different thresholds to determine the extent of conformational change were applied, with the results providing insight into the intrinsic flexibilities of amino acid side chains, and the extent of side-chain conformational change on ligand binding. Particular emphasis will be placed on complexes of HIV-1 Protease, Endothiapepsin, and Streptavidin. The implications of these results for the prediction of side-chain conformation in the context of rational drug design will be discussed.

COMP 20 [829940]: Small-system effects in molecular dynamics simulations

Randall B. Shirts, Scott R. Burt, and Aaron M. Johnson, Department of Chemistry and Biochemistry, Brigham Young University, C100 Benson Building, Provo, UT 84602, Fax: 801-422-0153, randy_shirts@byu.edu

Molecular dynamics simulations are now ubiquitous for predicting bulk behavior of chemical systems. Applications range from drug discovery to materials engineering. I present some fundamental results for finite-size systems that indicate the nature and size of corrections that must be considered in extrapolating finite-system results to bulk properties. These corrections include (1) the difference in effective temperature between particles of different mass induced by the center of mass constraint, (2) corrections to the velocity and kinetic energy distributions and associated statistics, (3) corrections to equations of state like the virial expansion, and (4) corrections to relations depending on the average and average relative velocity (e.g. mean free path and collision rate). In most cases, it is more efficient to apply corrections to results from a small simulation than take large enough systems to make the corrections negligible.

COMP 21 [830710]: Sub-microsecond conformer transitions of protein inhibitor, fasciculin to acetylcholinesterase

Jennifer Bui, Department of Chemistry/ Biochemistry, University of California San Diego, 9500 Gilman Drive MC 0365, la jolla, CA 92093, Fax: 858-534-7042, jbui@mccammon.ucsd.edu, and J Andrew McCammon, Howard Hughes Med. Inst., NSF Ctr. Theor. Biol. Physics, Dept. Chem. and Biochem., and Dept. Pharmacol, University of California, San Diego

Protein-protein interactions are ubiquitous in biological system. Recognition of binding is a key to high affinity interactions of protein complexes. The fasciculins (FAS), 61-amino-acid peptides, are potent inhibitors of synaptic acetylcholinesterase. Four fasciculins have been characterized to date: FAS1 and FAS2 from the venom of Dendroaspis angusticeps, toxin C from the venom of D.polylepis and FAS3 from the venom of D.viridis . The sequences of FAS1 and FAS2 are nearly identical and differ only by one residue at the position 47. The dynamic nature of the encounter between FAS molecules and acetylcholinesterase can shed light on conformational variations before and after bindings. This work presents the dynamical study, in terms of the accessible conformational space, at sub-microsecond time scale. Employing principle component analysis and clustering analysis methods in analyzing the sub-microsecond molecular dynamics (MD) trajectories of FAS1 and FAS2, and compared with a 15-nanosecond MD simulation of FAS bound to acetylcholinesterase, the important modes of conformational variations and transitions of FAS upon complexation have been identified and will be discussed.

COMP 22 [833736]: A gating mechanism proposed from a 15 nanosecond simulation of a complete human alpha-7 nicotinic acetylcholine receptor model

Richard James Law, Chemistry & Biochemistry, UCSD/HHMI, mc0365, 9500 Gilman Drive, La Jolla, CA 92093, Fax: 858-534-4974, rlaw@mccammon.ucsd.edu, J. Andrew McCammon, Howard Hughes Medical Institute, Department of Chemistry and Biochemistry, University of California, San Diego, and Richard Henchman, Department of Chemistry and Biochemistry, University of California, San Diego

The nicotinic acetylcholine receptor (nAChR) is a well characterized ligand gated ion channel yet a proper description of the mechanisms involved in gating, opening, closing, ligand binding, and desensitization does not exist. Until recently, atomic resolution structural information on the protein was limited, but with the production of the X-ray crystal structure of the L.stagnalis acetylcholine binding protein (AChBP) and the electron microscopy (EM) image of the transmembrane domain of the torpedo electric ray nicotinic channel, we were provided with a window to examine the mechanism by which this channel operates. A 15 nanosecond (ns) all-atom simulation of a homology model of the homomeric human α7 form of the receptor was conducted, in a solvated POPC bilayer, and examined in detail. The receptor was unliganded. The structure undergoes a twist-to-close motion that correlates movements of the C-loop in the ligand binding domain, via the β10-strand that connects the two, with the 10� rotation and inward movement of two non-adjacent subunits. The Cys-loop appears to act as a stator around which the α-helical transmembrane domain can pivot and rotate relative to the rigid β-sheet binding domain. The M2-M3 loop may have a role in controlling the extent or kinetics of these relative movements. All of this motion, along with essential dynamics analysis, is suggestive of the direction of larger motions involved in gating of the channel.

COMP 23 [820326]: A comparison of MacroModel methodologies for ligand conformation generation

John C. Shelley, Schrodinger Inc, 1500 SW First Avenue, Suite 1180, Portland, OR 97201, Fax: 503-299-4532, jshelley@schrodinger.com

Generation and consideration of collections of ligand conformations is often an important part of computational drug design efforts. The type of generation employed is often influenced by the time available, number of ligands to search, desired thoroughness of the search, and sensitivity to the quality of the generated conformations. We present the results for a number of searching approaches using MacroModel including random torsional searches (MCMM), low mode searches, mixed mode searches and systematic searches applied to a large collection of molecules. The systematic searches are carried out with a new MacroModel module, ConfGen, which is designed to generate quality ensembles of ligand conformations rapidly. Quality of the results will be judged by identification of the global minimum structure, having a conformer with a small RMS relative to the same ligand from crystal structures of protein-ligand complexes and completeness of the conformational collections generated.

COMP 24 [821067]: Low-coordination silicon compounds. Interplay and synergism between experiment and theory

Yitzhak Apeloig, Department of Chemistry and The Lise Meitner – Minerva Center for Computational Quantum Chemistry, Technion – Israel Institute of Technology, Haifa 32000, Israel, Fax: 972-4-8292000, chrapel@tx.technion.ac.il

Silicon is the closest congener of carbon. Yet the fundamental properties of many silicon and carbon compounds are very different. This is especially evident for low – coordination compounds, such as multiple bonds to silicon or silylenes whose chemistry has begun to be unraveled only in the last two decades, following the synthesis of the first stable compounds of these types. These exciting developments were occurring at the time when computational methods reached maturity, and consequently quantum mechanical calculations made numerous crucial contributions to silicon chemistry, and in many cases theoretical predictions preceded and directed experimental work.

Some of the intriguing differences between silicon and carbon compounds will be discussed, demonstrating how the synergism between theory and experiment can be used to discover new chemistry. Three examples are:

(a)Carbenes dimerize to give double bonds while silylenes (R₂Si:) in addition to disilene 1, can also produce novel low – coordination cyclic dimers of type 2, unprecedented in organic chemistry. The factors which control the dimerization mechanism of silylenes will be discussed.

(b) The surprising structures of multiply-bonded silicon compounds in general, and trisilaallene in particular will be discussed.

(c) The first compound with a Si≡Si triple bond was recently isolated. The unique nature of triple bonds to silicon and the interplay of theory and experiment in the quest of their synthesis will be discussed.

COMP 25 [818848]: The schizophrenic effect of geminal fluorination on the kinetic stability of molecules containing strained rings

Weston Thatcher Borden¹, David A. Hrovat¹, Christine Isborn², Scott B. Lewis³, and Stephen Getty⁴. (1) Department of Chemistry, University of North Texas, P.O. Box 305070, Denton, TX 76203-5070, Borden@unt.edu, (2) Department of Chemistry, University of Washington, (3) Department of Chemistry, James Madison University, (4) Du Pont Chemical Company

Experiments by Dolbier and coworkers have found that geminal fluorination has a large effect on lowering the barrier to cis-trans isomerization of 1,2-dimethylcyclopropane and an even larger effect on lowering the barrier to racemization of optically active 1-ethyl-2-methylcyclopropane. In contrast, Dolbier has found that that geminal fluorination has almost no effect on the barrier to rearrangement of methylenecyclopropane. Finally, Lemal and coworkers have shown that geminal fluorination dramatically stabilizes [2.2.2]propellane toward ring opening to 1,4-dimethylenecyclohexane. In order to understand these very different effects of geminal fluorination, ab initio calculations have been performed; and the results of these calculations will be discussed.

COMP 26 [804177]: Stable carbon(0) compounds: Theoretical analysis of molecules with unusual bonding situations

Gernot Frenking, Fachbereich Chemie, Philipps Universit�t Marburg, Hans-Meerwein-Strasse, D-35039 Marburg, Germany, frenking@chemie.uni-marburg.de

We report about a combined theoretical/experimental study of the structures and bonding situation of carbodiphosphoranes C(PR₃)₂. The results suggest that carbodiphosphoranes should be considered as Lewis acid-base complexes between a naked carbon atom with the electron configuration (2s)²;(2p_z)²;(2p_x)⁰;(2p_y)⁰ which serves as the acceptor moiety and two phosphane donors PR₃. Carbodiphosphoranes have two electron lone-pairs at the carbon atom which make C(PR₃)₂ a very strong Lewis base. This becomes evident by the fact that carbodiphosphoranes may even bind CO₂ in a donor-acceptor complex which is stable enough to become the subject of an x-ray structure analysis.

COMP 27 [818555]: What IS the acidity of imidazole? An assessment of theoretical protocols for calculation of pKa values

Brian F. Yates and Alison M. Magill, School of Chemistry, University of Tasmania, Private Bag 75, Hobart TAS 7001, Australia, Fax: +61 3 6226-2858, Brian.Yates@utas.edu.au

Recently we have exploited the method of Liptak and Shields in calculating highly accurate pKa values for a number of heterocyclic systems. However in attempting to carry out a series of systematic benchmark calculations on imidazole, we discovered that there is a significant discrepancy between high-level theoretical calculations and the accepted experimental value for the acidity of imidazole. We have studied this discrepancy further and carefully evaluated the various steps in the pKa calculation in an effort to determine the source of the disagreement with experiment.

COMP 28 [811535]: Intramolecular hydrogen bonds versus other weak interactions

Otilia M�, Departamento de Qu�mica, C-9, Universidad Aut�noma de Madrid, Cantoblanco, 28049-Madrid, Spain, Fax: +34-91-497-5238, otilia.mo@uam.es

The equilibrium geometries of β-chalcogenovinylaldehydes, HC(=X)-CH=CH-CYH (X = O, S; Y = Se, Te) correspond to chelated structures which are stabilized through the formation of X-H•••Y, X••••H-Y intramolecular hydrogen bonds or through X•••Y chalcogen-chalcogen interactions. The nature and strength of these interactions have been investigated through the use of high-level ab initio calculations, which show that for tellurium containing compounds the X••••TeH chalcogen-chalcogen interactions are much stronger that the X-H•••Te, X••••H-Te intramolecular hydrogen bonds, while the reverse stability order is observed for O and S containing compounds. The possible role of resonance assisted phenomena on the stability of these chelated structures was also investigated through a detailed analysis of the NMR properties of oxygen-containing systems. The conclusion was that neither the coupling constants nor the proton chemical shifts provide any evidence for the existence of resonance assisted phenomena.

COMP 29 [826389]: On the roles of π-stacking, hydrogen bonding and donor-acceptor interaction in acrolein complexes of chiral oxazaborolidinone catalysts

Ming Wah Wong, Department of Chemistry, National University of Singapore, 3 Science Drive 3, Singapore 117543, Singapore, Fax: +65-67791691, chmwmw@nus.edu.sg

Chiral N-sulphonylated 1,3,2-oxazaborolidin-5-ones have been shown to be efficient catalysts for asymmetric reactions. For instance, Corey et al have shown that N-tosyltryptophan-derived oxazaborolidinone catalyst is able to achieve an unprecedented enantioselectivity of greater than 200:1 for the Diels-Alder reactions between cyclopentadiene and 2-bromoacrolein. A better understanding of the origin of enantioselection is crucial to the rational development of new synthetic methodology and to the successful application and extension of enantioselective reactions. In this study, we have examined several acrolein complexes of Lewis acidic chiral N-sulphonylated oxazaborolidinones by means of ab initio and density functional calculations. The structures of these acrolein complexes indicate that π-stacking, hydrogen bonding and donor-acceptor interaction play essential roles in understanding the enantioselectivity of these chiral catalysts. We propose an alternative explanation to the high enantioselectivity of the oxazaborolidinone catalysts in terms of the most stable tridentate complex involves the s-trans acrolein. Furthermore, we find that the formyl hydrogen of acrolein prefers a hydrogen bond with the oxygen atom of the N-sulphonylated group in the acrolein-catalyst complexes and in the transition states. This is in sharp contrast to Corey's transition state model which favours the formation of formyl C-H•••O hydrogen bond with the ring oxygen.

COMP 30 [814789]: Dynamics on the way to forming glass

David Chandler, Department of Chemisty, 210 Gilman Hall, University of California, Berkeley, Berkeley, CA 94720, Fax: 510-643-1566, chandler@cchem.berkeley.edu, and Juan P. Garrahan, University of Nottingham

We have developed a new view of glass forming materials based upon a statistical mechanics of trajectory space. This space-time view has provided a geometric explanation of dynamic heterogeneity, and a series of distinguishing predictions, including dynamic scaling and universality. A few of our most recent predictions will be discussed.

COMP 31 [831533]: Structure and interactions in nematic-colloid dispersions

Tatiana G. Sokolovska, Ruslan O. Sokolovskii, and Gren N. Patey, Department of Chemistry, University of British Columbia, Vancouver, BC V6T1Z1, Canada, Fax: 604-822-2847, tata@chem.ubc.ca, patey@chem.ubc.ca

Microscopic theory and computer simulations are used to obtain molecular density and orientational distributions about colloidal particles immersed in a model nematic fluid. In the presence of a weak external field that serves to fix the bulk director, it is shown that these distributions are highly directional, and lead to strong, directional interactions amongst the colloidal particles. The nature and physical origin of the colloid-nematic structure, together with its dependence on the colloid-molecule interactions and on the applied field strength will be discussed.

COMP 32 [825909]: Simulation studies of structure and retention in chromatographic systems

J. Ilja Siepmann¹, Ling Zhang², Li Sun², Collin D. Wick², and Mark R. Schure³. (1) Departments of Chemistry, Chemical Engineering and Materials Science, University of Minnesota, 207 Pleasant St. SE, Minneapolis, MN 55455, Fax: 612-626-7541, siepmann@chem.umn.edu, (2) Department of Chemistry, University of Minnesota, (3) Theoretical Separation Science Laboratory, Rohm and Haas Company

Various driving forces have been suggested to explain retention in reversed-phase liquid chromatography (RPLC). To provide molecular-level information on structure and retention mechanism in RPLC, configurational-bias Monte Carlo simulations in the Gibbs and isobaric-isothermal ensembles were carried out for the following model systems/processes: (i) partitioning of solutes between a bulk n-hexadecane and water/methanol mixtures representing the retentive and mobile phase, respectively; (ii) an isolated n-octadecane chain solvated in water/methanol or water/acetonitrile mixtures; (iii) structure and retention of a model stationary phase consisting of dimethyl octadecyl silane chemisorbed on the (111) face of beta-cristobalite.

COMP 33 [848106]: Nucleation phenomena in polymer crystallization

Richard H. Gee, Chemistry and Material Science Directorate, Lawrence Livermore National Laboratory, Mail Code L-268, 7000 East Avenue, Livermore, CA 94550, Fax: 925-422-6810, gee10@llnl.gov

COMP 34 [833463]: Crossover from free Rouse to entangled chain dynamics in polyethylene melts

Jean-Paul Ryckaert, Polymer Physics, Universit� Libre de Bruxelles, CP223, boulevard de Triomphe, 1050 Bruxelles, Belgium, Fax: 02-650-56-75, jryckear@ulb.ac.be

Well equilibrated strongly entangled C1000 polyethylene melts in the 400K-500K temperature range are studied by molecular dynamics on the basis of an atomistic potential over a time window going from a few fs up to 20ns. This wide time range allows to extract from MD trajectories experimentally relevant time correlation functions, e.g. the incoherent intermediate scattering function Fs(Q,t) probed by neutron spectroscopy and the C-H bond orientational relaxation function G(t), probed by 13C NMR T1 relaxation measurements. These functions show successively an initial microscopic regime, a free Rouse regime and an entangled Rouse regime. Temperature effects in the dynamical relaxation will also be discussed as the dynamical interpretation of NMR data requires a chain relaxation model including the temperature dependence of the relevant relaxation times.

COMP 35 [816276]: Simulating polymersomes: A fruitful collaboration in rational coarse graining

Dennis E. Discher, Department of Chemical and Biomolecular Engineering, University of Pennsylvania, Philadelphia, PA 19104, discher@seas.upenn.edu, and Michael L. Klein, Department of Chemistry, Center for Molecular Modeling, University of Pennsylvania

Block-copolymer amphiphiles have been observed to assemble into vesicles and other morphologies long known for lipids but with remarkably different properties. In a fruitful collaboration making use of phenomolgical measures from the Discher group, coarse-grain molecular dynamics (CG-MD) developed by the Klein group have been extended to elaborate structures and properties of diblock copolymer assemblies in water. By varying the hydrophilic/hydrophobic ratio of the copolymer in line with experiment, bilayer, cylindrical and spherical micelle morphologies spontaneously assemble. Varying the molecular weight (MW) with hydrophilic/hydrophobic ratio appropriate to a bilayer yields a hydrophobic core thickness that scales for large MW as a random coil polymer, in agreement with experiment. The extent of hydrophobicsegment overlap in the core increases nonlinearly with MW, indicative of chain entanglements and consistent with the dramatic decrease reported for lateral mobility in polymer vesicles. Calculated trends with MW as well as hydrophilic/hydrophobic ratio thus agree with experiment, demonstrating that CG-MD simulations provide a rational design tool for novel Materials Science with block copolymers and their assemblies.

COMP 36 [819363]: Protein-ligand interactions and computer-aided drug design

William L. Jorgensen, Juliana Ruiz-Caro, and Julian Tirado-Rives, Department of Chemistry, Yale University, New Haven, CT 06520-8107, Fax: 203-432-6299, william.jorgensen@yale.edu

Drug development is being pursued through computer-aided design, synthesis, and assaying. The design begins with use of the BOMB program, which rapidly constructs combinatorial libraries given the structure of the target protein and a selected core and substituents. BOMB grows the analogs inside the protein's binding site, performs a thorough conformational search, and estimates the analog's binding affinity or activity using scoring functions. The QikProp program is applied to filter all designed molecules to insure that they have drug-like properties including solubility and cell permeability. MC/FEP simulations are then performed to refine the predictions for the best scoring leads using hundreds of explicit water molecules and extensive sampling for the protein and ligand. Recent methodological advances and representative applications will be presented with emphasis on inhibitor development for HIV reverse transcriptase.

Some references: Prediction of Drug Solubility from Structure. W. L. Jorgensen and E. M. Duffy, Adv. Drug Delivery Reviews, 54, 355-366 (2002). Validation of a Model for the Complex of HIV-1 Reverse Transcriptase with the Novel Non-nucleoside Inhibitor TMC125. Blagovi�, M. U.; Tirado-Rives, J.; Jorgensen, W. L., J. Am. Chem. Soc., 125, 6016-6017 (2003). General Model for Estimation of the Inhibition of Protein Kinases Using Monte Carlo Simulations. Tominaga, Y. & Jorgensen, W. L., J. Med. Chem. 47, 2534-2549 (2004). The Many Roles of Computation in Drug Discovery. Jorgensen, W. L., Science 303, 1813-1818 (2004).

COMP 37 [807610]: Free energy calculations in structure-based drug design

J Andrew McCammon, Howard Hughes Med. Inst., NSF Ctr. Theor. Biol. Physics, Dept. Chem. and Biochem., and Dept. Pharmacol, University of California, San Diego, 9500 Gilman Drive, MC0365, La Jolla, CA 92093-0365, Fax: 858-534-4974, jmccammon@ucsd.edu

Quantitative knowledge of binding affinities and selectivities based on standard free energies can be helpful in the engineering of new drugs. This talk will describe recent progress in the methods for computing such free energies of binding.

Images and animations related to this work can be found at the website http://mccammon.ucsd.edu/

COMP 38 [824444]: Calculation of binding affinities

Michael K. Gilson, Center for Advanced Research in Biotechnology, University of Maryland Biotechnology Institute, 9600 Gudelsky Drive, Rockville, MD 20850, Fax: 301-738-6255

Abstract text not available.

COMP 39 [834529]: Challenges in pKa calculations for internal residues in proteins

Bertrand Garcia-Moreno, Daniel G. Isom, and Carolyn A. Fitch, Department of Biophysics, Johns Hopkins University, 3400 N. Charles St., Baltimore, MD 21218, Fax: 410-516-4118, bertrand@jhu.edu

Internal ionizable residues in proteins play significant biological roles in all aspects of energy transduction and catalysis. They titrate with shifted pKa values because the protein is not as polarizable as water. They are extremely difficult to reproduce with structure-based calculations because the dielectric response of proteins is difficult to treat quantitatively. A large set of apparent pKa values of internal Lys and Glu residues in staphylococcal nuclease measured recently in our lab suggest that the protein interior is considerably more polarizable than is generally acknowledged. The molecular origins of this high apparent polarizability are not yet known. We will present data to show that quantitative treatment of the reorganization of the protein concomitant with ionization of internal groups is the most serious challenge faced by computational methods for calculation of electrostatic energies in environments sequestered from bulk solvent (i.e. protein interior, active sites, and interfaces between macromolecules).

COMP 40 [817575]: Are fragment-protein binding energies sufficient to predict compound affinity? p38 as a case study in structure based design using fragments

Frank P. Hollinger¹, Zenon Konteatis¹, Enrique L. Michelotti², Ted T. Fujimoto¹, Jennifer L. Ludington¹, Michael Karpusas³, Marina Bukhtiyarova³, Mike Saporito³, Xiaomei Chai³, Katrina Northrop³, and Eric Springman³. (1) Technology & Informatics, Computational Chemistry, Locus Pharmaceuticals, Inc, Four Valley Square, 512 Township Line Road, Blue Bell, PA 19422, Fax: 215-358-2020, fhollinger@locuspharma.com, (2) Department of Chemistry, Locus Pharmaceuticals, Inc, (3) Department of Biology, Locus Pharmaceuticals, Inc

The development of a novel grand canonical Monte Carlo (GCMC) simulation paradigm permits the generation of binding free energies for fragments to a protein. The output of such a simulation provides all the information necessary to accurately identify high affinity interaction sites, binding sites and druggable binding sites on a protein surface. Using novel analysis tools we are able to combine fragments into synthetically accessible drug-like molecules for evaluation against a desired protein target. A case study will be presented describing the design of novel, potent and selective allosteric inhibitors for p38-alpha. Biological and structural experimental data will be presented which validates our process of using fragment based ligand design approaches.

COMP 41 [821623]: Computational studies of protein-solvent and protein-ligand interactions

Kaushik Raha and Kenneth M. Merz Jr, Department of Chemistry, The Pennsylvania State University, 105 Chemistry Building, University Park, PA 16802, kxr205@psu.edu

We have studied electrostatic properties of proteins and their interactions with solvent and ligands at various levels of theory using computational methods such as molecular dynamics simulations, molecular mechanics, and linear-scaling quantum mechanics based methods. We elucidate dielectric permittivity of proteins and find that it is a unique property that depends on the charge, composition and conformational flexibility of the protein. We have also designed a quantum mechanics based scoring function to calculate free energy of binding in protein-ligand interaction that can capture trends in diverse sets of protein-ligand complexes. We also show the importance of electrostatic interactions in discriminating native binding mode from decoy “poses”. In related studies we use pairwise decomposition of residue interaction energy to dissect the interaction between different parts of the protein and the bound ligand. Finally, we demonstrate that these methods can be used in structure based drug discovery to understand structure-activity relationships.

COMP 42 [843431]: Can topological indices transmit information on properties but not on structures?

Alexandru T. Balaban, Marine Sciences Department, Texas A&M University at Galveston, 5007 Avenue U, Galveston, TX 77551, Fax: 409-740-4787, balabana@tamug.edu

Information on atom connectivity of hydrogen-depleted formulas for organic compounds can be provided by various topological indices, which use different graph invariants for this purpose. If one wishes to include information on stereochemistry, more sophisticated approaches are needed. Topological indices are based on local (vertex) graph invariants, LOVIs, which can be integers or real numbers. For a given structure, these LOVIs are assembled into one numerical value (the topological index, TI) which again can be an integer or a real number. So far, no TI was designed which would be able to be in a one-to-one correspondence with a chemical structure; there are always several structures that correspond to a numerical value of a TI, giving rise to the degeneracy of TIs. The higher the degeneracy, the lower the discrimination ability of the TI. However, real-number LOVIs and real-number TIs have lower degeneracy than integer-number analogs. For such analogs and a few real-number TIs with low discrimination ability, there exist algorithms for finding all possible structures associated with a given TI; this is the so-called “inverse problem” that arises when one wishes to test structures fitting in a QSAR, QSTR or QSPR “window” or interval of biological activity, toxicity, or value of a physical property, respectively. A brief review will be presented for TIs of high and low degeneracy, for TIs that have algorithms for the corresponding inverse problems, and for TIs which offer promise that they could be associated with biological, physical or chemical properties without the possibility of retrieving information on the chemical constitution of the corresponding compounds.

COMP 43 [851394]: Coding and decoding chemical structure information

Johann Gasteiger and Dimitar Hristozov, Department of Organic Chemistry, Computer-Chemie-Centrum, University of Erlangen-Nuremberg, Naegelsbachstrasse 25, 91052 Erlangen, Germany, Fax: +49-9131.8526566, gasteiger@chemie.uni-erlangen.de

Chemical structures can be represented to various degrees of sophistication, from the constitution, through the 3D structure, to molecular surfaces. At each level, different physicochemical properties can be considered. Mathematical transformations allow one to obtain uniform structure codes for each different level of structure representation. These structure codes can be used to model a variety of physical, chemical or biological properties. Studies on the WOMBAT database will be reported. It will also be shown to what extent structure information can be regained from these structure codes.

COMP 44 [821669]: Molecular shape and electrostatics in the encoding of relevant chemical information

Anthony Nicholls, OpenEye Scientific Software Inc, 3600 Cerrillos Rd., Suite 1107, Santa Fe, NM 87505, anthony@eyesopen.com, and J. Andrew Grant, Lead Discovery, AstraZeneca Pharmaceuticals Ltd

Molecular shape and electostatic profile are key descriptors of molecular activity. In addition, they derive from the underlying chemical composition but not uniquely, i.e. different molecules may have similar shape and electrostatics. As such they have the property of a searchable one-way hash function: they encode much that is relevant of the underlying molecule while hiding its identity. This presentation will discuss various methods by which shape and electrostatics may be represented for rapid search and retrieval and used as a "safe" surrogate for chemical content.

COMP 45 [852042]: Confusing descriptors: Where chemical information gets dizzy

Cristian G. Bologa, Marius Olah, and Tudor I. Oprea, Division of Biocomputing, University of New Mexico School of Medicine, MSC 084560, 1 University of New Mexico, Albuquerque, NM 87131-0001, toprea@salud.unm.edu

Structures from WOMBAT (WOrld of Molecular BioAcTivity) [1] were investigated with several descriptor systems. For 79,483 unique non-stereoisomeric compounds, we found multiple “confused“ instances: For 2D-descriptors, 314 duplicates (0.4%) across 80 descriptors (487 pairs); for MESA-implemented [2] MDL keys, 4391 duplicates (5.5%) across 320 keys; for Daylight fingerprints [3], 7166 duplicates (9.0%) for 512-keys, 5010 duplicates (6.3%) for 1024-keys, and 4092 duplicates (5.1%) at the 2048 level. The WOMBAT-derived set of 512 keys had 6202 (7.8%) duplicates. Our results indicate that, for several chemical descriptor systems, it is not always possible to provide a 1:1 map between chemical structure and chemical description. This implies that we can devise an information–rich, yet “confused” descriptor system, i.e., a chemical information exchange tool allowing for chemical structure ambiguity.

[1] WOMBAT is available from http://www.sunsetmolecular.com [2] The MDL 320 keys fingerprinter is available from http://www.mesaac.com [3] The Daylight fingerprinter is available from http://www.daylight.com

COMP 46 [804198]: Anonymous sd (.asd) files

Tim Clark, Friedrich-Alexander-Universite Erlangen-N�rnberg, Computer-Chemie-Centrum, Nagelsbachstrasse 25, D-91052 Erlangen, Germany, Fax: +49-9131-8526565, clark@chemie.uni-erlangen.de

A complete description of molecules and their intermolecular binding properties is presented that does not include information about the 2D structure. The molecule is described in terms of its shape (as an shrink-wrap isodensity surface) and the values of four local properties (the molecular electrostatic potential and the local ionization energy, electron affinity and polarizability) at the triangulation points on this surface. All five descriptors (the shape and the four local properties) are fitted to spherical-harmonic expansions, whose degree can be varied in order to adjust the resolution of the molecular description. The anonymous sd-file (.asd) contains only the coefficients of the five spherical-harmonic fits and thus provides an extremely information-rich description of the molecule without its 2D structure.

COMP 47 [852038]: Similarity-based descriptors (SIBAR) as tool for exchange of chemical information

Gerhard F. Ecker, Barbara Zdrazil, and Dominik Kaiser, Department of Pharmaceutical Chemistry, University of Vienna, Althanstrasse 14, Vienna A-1090, Austria, Fax: +431-4277-9551, gerhard.f.ecker@univie.ac.at

Recently we published the successful application of a set of new descriptors based on similarity values, denoted as SIBAR-descriptors (Similarity Based SAR). These descriptors are based on calculation of similarity (on basis of euclidian distances) for each compound of the data set to each compound of a reference set, using common descriptors. These euclidian distances (= similarity values) are then further used for QSAR-studies. Both the references set as well as the descriptors used for calculating the SIBAR-values are tailored to the specific QSAR-problem. Best results have been obtained when targeting ADMET-problems. In any case it needs the knowledge of the reference set to retrieve the corresponding descriptors. Assuming that only the descriptors for calculating the SIBAR-values, but not the structures of the reference compounds are available, it should be impossible to trace back the chemical structure of the original compounds of the training set.

COMP 48 [851558]: Encoding and exchange of chemical information using substructural molecular fragments

Alexandre Varnek, Denis Fourches, and Vitaly P. Solov’ev, Laboratoire d’Infochimie, Louis Pasteur University, 4, rue B. Pascal, Strasbourg 67000, France, Fax: +33-3-88416104, varnek@chimie.u-strasbg.fr

In this presentation, we describe how chemical information can be encoded in substructural molecular fragments, then used for “in silico” design of new compounds. The Substructural Molecular Fragments method is based on the representation of a molecule by its fragments and on the calculation of their contributions to a given property. Two different classes of fragments are considered “sequences” and “augmented atoms”. The sequences represent the shortest path between each pair of atoms; their length vary from 2 to 15 atoms An augmented atom represents a selected atom with its first coordination sphere. The both classes of fragments involve either atoms and bonds, or atoms only, or bonds only. Once a given compound is split into constitutive fragments, any its quantitative property is calculated from the fragments contributions using several linear and non-linear fitting equations. The best structure-property models are selected according to statistical criteria. Thus, the information concerning a given data set is stored in the files containing types of fragments, their contributions and corresponding fitting equations. In the framework of the ISIDA project (http://infochim.u-strasbg.fr/recherche/isida/index.php) we have developed a knowledgebase which stores the structure-property models based on fragment descriptors using PostgreSQL environment. Since the model is loaded to the knowledgebase, it immediately becomes available for all users of INTRANET via a client application. The stored models can be efficiently used for a virtual screening of large combinatorial libraries. Several examples of application of chemical information encoded in substructural molecular fragments for “in silico” design of new compounds possessing desirable chemical or biological activities will be given.

COMP 49 [810564]: One- and two-bond spin-spin coupling constants across X-H-Y hydrogen bonds

Janet E. Del Bene, Department of Chemistry, Youngstown State University, One University Plaza, Youngstown, OH 44555, Fax: 330-941-1579, jedelbene@ysu.edu

One-bond X-H (¹J_X-H) and H-Y (^1hJ_H-Y) and two-bond X-Y (^2hJ_X-Y) spin-spin coupling constants across X-H-Y hydrogen bonds have been computed for X and Y the second-period elements ¹³C, ¹⁵N, ¹⁷O, and ¹⁹F, using the equation-of-motion coupled cluster singles and doubles method (EOM-CCSD). Relationships have been established among the signs and magnitudes of coupling constants, X-Y and X-H distances, and hydrogen bond type. For complexes with traditional hydrogen bonds, the reduced Fermi-contact terms and the reduced spin-spin coupling constants ¹K_X-H and ^2hK_X-Y are positive, except for ^2hK_F-F for the equilibrium structure of (HF)₂, while ^1hK_H-Y is negative_.As the degree of proton-shared character of the hydrogen bond increases, all reduced coupling constants become positive. The signs of ¹K_X-H, ^1hK_H-Y, and ^2hK_X-Y are interpreted in terms of the Nuclear Magnetic Resonance Triplet Wavefunction Model (NMRTWM). Determination of the signs and magnitudes of coupling constants could be useful for confirming the presence or absence of a proton-shared hydrogen bond.

COMP 50 [808708]: Theoretical modeling of single molecule magnets

E. R. Davidson, Department of Chemistry, University of Washington, Seattle, WA 98195-1700, erdavid@u.washington.edu

Single molecule magnets usually have several transition metal centers with high spin. These spins couple to form molecules with net magnetic moments. In some cases the zero field splitting leads to molecules that act like magnets, albeit with a Curie temperature below 5 K. Traditionally such molecules are described by a Heisenberg Hamiltonian that assumes a net spin for each metal center and an empirical exchange coupling between centers. Methods for extracting the spin and exchange parameters from calculations of the wave function and energy will be discussed. Reasons for not using conventional quantum chemistry methods such as UHF will be explained.

COMP 51 [812256]: Donor/acceptor interaction: Electronic structural analysis and associated vibronic features

Marshall D. Newton, Department of Chemistry, Brookhaven National Laboratory, Building 815, Upton, NY 11973, Fax: 631-344-5815, Newton@bnl.gov

Electronic interaction between localized donor and acceptor sites, especially as mediated by intervening molecular spacers (‘bridges') are fundamental to many features of chemical behavior, including long-range electron ( and hole) transfer. The analysis of such interactions and schemes for computational implementation are discussed, including the role of nuclear modes in modulating the electronic structure effects. Results for specific molecular systems are illustrated.

COMP 52 [817566]: Correlation potential in density functional theory and recollections of Pople's entry into DFT

Mel Levy, Departments of Physics and Chemistry, North Carolina A & T State University/Greensboro, NC and Tulane University/New Orleans, Louisiana, 906 Haddington Court, Whitsett, NC 27377, mlevy@tulane.edu

The first part of the lecture will concern interactions with Professor Pople in 1990, when he began to think about DFT. At that time, I gave a talk at Carnegie Mellon University and he much later said that our discussions then influenced him into analyzing the possibilities of the theory for computational chemistry. I will also recall a special joyous occasion in New Orleans in 1999 and earlier help that he gave me. The second part of the lecture will discuss the generation of the DFT correlation potential both from the "density condition" within perturbation theory [1,2] and as an explicit functional of the density without the use of perturbation theory. Recent progress in time-independent excited-state DFT [3] will also be discussed.

1. A. Goerling and M. Levy, Int. J. Quantum Chem., Symp. 29, 93 (1995). 2. S. Ivanov and M. Levy, J. Chem. Phys. 116, 6924 (2002). 3. M. Levy and �. Nagy, “Variational Density-Functional Theory for an Individual Excited State”, Phys. Rev. Lett. 83, 4361 (1999).

COMP 53 [807156]: Recent progress in the development of exchange-correlation functionals

Gustavo E. Scuseria, Department of Chemistry, Rice University, Houston, TX 77005, guscus@rice.edu

This presentation will address our current efforts to develop more accurate exchange-correlation functionals for DFT. The functionals to be discussed include a new meta-GGA denoted TPSS [1], a screened exchange hybrid especially designed with solids in mind [2], local hybrids [3], and a current (j) dependent extension of PBE [4]. Extensive benchmarks and applications to actinide oxides (UO₂ and PuO₂) will also be presented.

[1] J. Tao, J. P. Perdew, V. N. Staroverov, and G. E. Scuseria, Phys. Rev. Lett. 91, 146401 (2003).

[2] J. Heyd, G. E. Scuseria, and M. Ernzerhof, J. Chem. Phys. 118, 8207 (2003).

[3] J. Jaramillo, M. Ernzerhof, and G. E. Scuseria, J. Chem. Phys. 118, 1068 (2003).

[4] S. N. Maximoff, M. Ernzerhof, and G. E. Scuseria, J. Chem. Phys. 120, 2105 (2004).

COMP 54 [817218]: A new hybrid DFT functional: Accurate description of excited states, charge-transfer states, and van der Waals interactions

Kimihiko Hirao, Department of Applied Chemistry, University of Tokyo, Tokyo, Japan, Fax: 81-3-5841-7241, hirao@qcl.t.u-tokyo.ac.jp

Density functional theory (DFT) has advanced to one of the most popular theoretical approaches to calculate molecular properties. The first-order molecular properties (energies, geometries, frequencies, dipole moments, etc) are well predicted by local GGA functionals. However, DFT fails to describe induced or response properties. Although the valence-excited states can be well described by time-dependent DFT (TDDFT), TDDFT significantly underestimates the Rydberg and charge transfer (CT) excitation energies. The computed oscillator strengths have substantial errors. Also DFT fails to describe van der Waals interactions. This failure has been attributed to the wrong long-range behavior of the standard exchange functionals. Recently we have proposed a long-range exchange correction scheme for GGA functional. In the scheme, the two-electron operator is separated into the short-range and long-range parts by using the standard error functions. The long-range exchange interaction is described by the Hartree-Fock exchange integral and the short-range part is replaced by the GGA exchange functional. The present long-range corrected functional has been successfully applied to the various molecular properties. It remedies the underestimation of Rydberg excitation energies and reproduces the correct asymptotic behavior of the CT excitation energies. The van der Waals interactions can be described accurately by the present scheme with combining Andersson's potential.

COMP 55 [820956]: Molecular dynamics simulations of the open and closed conformations of KirBac3.1.: Structural changes during ion channel gating

Carmen Domene, Physical and Theoretical Chemistry Laboratory, Department of Chemistry, University of Oxford, South Parks Road, Oxford OX1 3QZ, United Kingdom, Fax: 44-1865275410, carmen.domene@chem.ox.ac.uk

A central challenge remaining in ion channel biophysics is to understand the mechanisms of channel gating. Ion channels do not stay open all of the time. Instead, they are 'gated' by either the binding of small molecules e.g. neurotransmitters or intracellular regulators to the channel protein or by changes in voltage across the membrane. This reversible transition from the closed to the open state is believed to operate via conformational changes.

Molecular Dynamics Simulations were used to test the validity of the open state model of the inward rectifying K+ channel KirBac3.1. This channel was experimentally captured in its closed and open states providing 'snapshots' of the gating process. For the first time, we were able to compare conformational changes associated with channel opening in the same family of K+ channels.

COMP 56 [832905]: Molecular modeling of ion channels: From ab initio calculations to structural bioinformatics

Paolo Carloni, Sector of statistical and biological physics and INFM-Democritos Center, International School for Advanced Studies, SISSA/ISAS, 34100 Trieste, Italy, Via Beirut 2-4, 34014 Trieste, Italy, Fax: 0039-040-3787-528, carloni@sissa.it
Abstract text not available.

COMP 57 [821124]: Activation of Shaker B, a voltage-gated potassium channel

Mounir Tarek and Werner Treptow, Equipe de Dynamique des Assemblages Membranaires, Universit� Henri Poincar�, UMR 7565 CNRS-UHP, BP 239, Vandoeuvre-L�s-Nancy Cedex 54506, France, Fax: 33-3-83 68 43 87, mtarek@edam.uhp-nancy.fr

Excitability is an electrical property of the membrane of excitable tissues, e.g., neurons, cardiac and muscular fibers. Voltage-gated (Kv) ion channels are complex molecular structures that control this cellular excitability at the molecular level. These proteins enable ions to flow through the plasmatic membrane in response to variations of the local electrical-chemical potential. Structural and functional characterization of such ion channels is crucial and has direct applications in pharmaceutical and medical research. In this paper, we present results from MD simulations of Shaker B, a voltage-gated potassium channel, and related channels aimed at investigating the stability, the activation mechanism, and the conductive properties of Kv channels.

COMP 58 [814284]: Mechanism of proteolysis of anthrax lethal factor. An ab initio and hybrid QM/MM molecular dynamics study

Alessandra Magistrato, INFM-Democritos Center and International School for Advanced Studies (SISSA/ISAS), via Beirut 2-4, Trieste 34014, Italy, Fax: +39-040-3787529, alema@sissa.it

The disease anthrax is caused by lethal factor (LF), an enzyme component of the toxin produced by the bacterium Bacillus Anthracis. Our studies are devoted to shed light on the binding and proteolytic mechanism of MAPKK kinase family promoted by anthrax lethal factor. At first, based on the X-ray structure of LF we have provided an understanding of the structural determinants and the hydrogen bond network that surrounds the LF active site, using static and dynamic density functional (DFT) calculations. Subsequently, classical molecular dynamics simulations have been performed on the entire protein structure and the solvent waters in order to clarify the binding and the specific substrate-protein interactions of MAPKK to the active site of LF. Finally, hybrid quantum/classical (QM/MM) molecular dynamics simulations have been performed on the entire protein structure and the solvent in order to understand the exact mechanism by which LF cleaves the NH₂-termini of the MAPK-kinase family.

COMP 59 [816047]: Structure and function of vanadium haloperoxidases

Simone Raugei, Statistical and Biological Physics, SISSA, Via Beirut 2-4, Trieste 34014, Italy, Fax: 0039 040 3787528

Since the discovery of vanadium-containing enzymes two decades ago, there has been a growing interest in the biological, pharmacological, and industrial applications of vanadium. Among these enzymes, haloperoxidases, which contain vanadium(V) as vanadate or related ions, are the most efficient halide oxidants known to date. Therefore, it is of interest to study these enzymes for their appealing utility in industrial-scale biocatalytic conversions. In the presence of hydrogen peroxide, this class of enzymes catalyzes the two-electron oxidation of a halide ion, X^-, to the hypohalous acids, HXO. HXO can further react with a wide range of nucleophilic molecules to yield a number of halogenated substances. By employing ab initio and QM/MM approaches, we investigated the electronic structure of the active site vanadate moiety, and the bonding pattern of the peroxo group and of halides anions. Furthermore, the formation of the intermediate peroxide form was also examined by QM/MM MD.

COMP 60 [822030]: Understanding protein-ligand interactions and binding free energy with an empirical solvation-based model

Glen E. Kellogg¹, Micaela Fornabaio¹, Pietro Cozzini², Andrea Mozzarelli², Francesca Spyrakis², and Donald J. Abraham¹. (1) Department of Medicinal Chemistry and Institute for Structural Biology & Drug Discovery, Virginia Commonwealth University, Box 980540, Richmond, VA 23298-0540, Fax: 804-827-3664, glen.kellogg@vcu.edu, (2) Department of Biochemistry and Molecular Biology, University of Parma

We have been developing a number of computational models and tools to explore the types and strengths of protein-ligand interactions. Our overarching goal is to build an understanding of the phenomena surrounding ligand binding in terms of simple and intuitive principles. Thus, we have presented HINT (Hydropathic INTeractions) as a uniquely empirical paradigm for free energy scoring of biomolecular interactions. The key to HINT is the experimental data from water/1-octanol partitioning of small molecules, an event not unlike ligand binding. We have shown that this model provides predictions with accuracy comparable to other free energy methods, and have recently extended the method to enhance our predictions by including: a) the effects of bridging water in the active site, and b) ionization states of protein residues and ligand functional groups in the active site (through computational titration). In this presentation we will report on a detailed examination of the XSCORE data set (230 protein-ligand complexes) using these tools to prepare corrected molecular models and evaluate the resulting protein-ligand interactions.

COMP 61 [833444]: Exploring protein ligand interactions: Binding free energies, docking and scoring

Charles L. Brooks III, Molecular Biology, TPC6, Scripps Research Institute, 10550 N. Torrey Pines Rd, La Jolla, CA 92037, brooks@scripps.edu

In this talk I will review recent progress from our group in the development of methods and protocols that explore the nature of protein-ligand binding landscapes. These include free energy calculations with atomically detailed models, where we examine the role of ligand configurational entropy loss and solvent expulsion in the binding process. I will also discuss key factors we have found that influence the accuracy of scoring functions in protein-ligand docking and describe ongoing large-scale “Internet computing” experiments using Predictor@home and the BOINC middleware environment to explore protein grid-based docking calculations.

COMP 62 [832805]: Computational modelling of inhibition and catalysis in malaria proteases

Johan Aqvist, Department of Cell and Molecular Biology, Uppsala University, Biomedical Center, POB 596, SE-751 24 Uppsala, Sweden, Fax: 46-18-536971, aqvist@xray.bmc.uu.se

The most virulent of the malaria parasites, P. falciparum, has several proteases involved in the degradation of host cell hemoglobin and these have emerged as highly promising drug targets. We will discuss recent theoretical and experimental work on inhibitor binding to the plasmepsin aspartic proteases, where methodological aspects of binding affinity calculations also will be addressed. Further, the so-called histo-aspartic protease (HAP) from P. falciparum presents a new type of active site and computer simulation studies of substrate binding and catalysis in this enzyme will be presented.

COMP 63 [821708]: Modeling protein-ligand interactions via high throughput docking and induced fit methods

Richard A. Friesner, Department of Chemistry, Columbia University, 3000 Broadway, MC 3110, New York, NY 10027, Fax: 212-854-7454

We will discuss application of high throughput docking methods (Glide) and protein structural refinement methods (Prime) to the problem of predicting the structures and binding affinities of protein-ligand complexes. We have examined a wide variety of pharmaceutically relevant test cases, including one system, p38 MAP kinase, for which we have obtained experimental binding affinities for selected ligands. A key aspect of our approach to the problem is separation of cases where the ligand fits into a given rigid receptor structure (in which case a rigid docking program should be able to correctly dock, and score, the ligand), from cases where the ligand is incompatible with the receptor conformation, i.e. the correct ligand pose would exhibit steric clashes with the conformation in question. In these latter cases, it is necessary to carry out an induced fit calculation, in which protein flexibility is incorporated. We have developed an automated induced fit protocol by combining Glide and Prime; this protocol will be described and examples of its use will be presented.

COMP 64 [846581]: Structure-based modeling of the hERG channel: A two-state linear interaction energy model for ligand binding

Brett A. Tounge, Computer Assisted Drug Discovery, Johnson & Johnson Pharmaceutical Research and Development, L.L.C, P. O. Box 776, Welsh and McKean Roads, Spring House, PA 19477-0776, Fax: 215-628-4985, btounge@prdus.jnj.com, Ramkumar Rajamani, Computer-Aided Drug Discovery, Johnson & Johnson Pharmaceutical R & D, and Charles H. Reynolds, Johnson & Johnson Pharmaceutical R&D
Homology models based on available K+ channel structures have been used to construct a two-state representation of the hERG cardiac K+ channel. These states are used to capture the flexibility of the channel. We show that this flexibility is essential in order to correctly model the binding affinity of a set of diverse ligands. Using this multiple state approach, a binding affinity model was constructed for set of known hERG channel binders. The predicted pIC50s are in good agreement with experiment (RMSD: 0.56 kcal/mol). In addition, these calculations provide structures for the bound ligands that are consistent with published mutation studies. These computed ligand bound complex structures can be used to guide synthesis of analogs with reduced hERG liability

COMP 65 [833356]: Hierarchical energy functions for the design of protein binding interfaces

David F. Green, Biological Engineering Division & Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, 32 Vassar St., Room 32-211, Cambridge, MA 02139, Fax: 617-252-1816, dfgreen@mit.edu, Michael D. Altman, Department of Chemistry & Computer Science and Artificial Intelligence Laboratory, MIT, and Bruce Tidor, Biological Engineering Division & Department of Electrical Engineering and Computer Science, MIT

Discrete conformational search methods have been demonstrated as a powerful means for designing novel proteins, and an obvious extension is to the design of protein complexes. Protein--protein binding interfaces pose a particular challenge due to the conformational flexiblity of both binding partners and the need to account for solvation. While the complexity of this search can be handled with various algorithms, simplified energetic and structural models are generally required. A hierachical approach, refining initial results with more accurate models, provides a means to search large spaces while ensuring accurate predictions. Here we outline several considerations for such a treatment. A framework for describing how ensembles of structures traverse the hierarchy is detailed. We also demonstrate how improvements at the lowest level affect the final results. While the methodology is detailed for designing protein complexes, it is easily extensible to related problems, including de novo ligand design and docking.

COMP 66 [826979]: Electronic structure studies of materials chemistry using embedded cluster models

Krishnan Raghavachari, Department of Chemistry, Indiana University, 800 E. Kirkwood Avenue, Bloomington, IN 47405, Fax: (812) 855-8300, kraghava@indiana.edu

Accurate quantum chemical calculations using embedded cluster models have been performed to investigate the surface chemistry of semiconductor materials and the endohedral chemistry of carbon nanotubes. In this talk, we consider new techniques for cluster termination as well as cluster embedding and discuss their applicability in such computational studies. Recent examples using our results to provide assignments and novel interpretations of experimental spectroscopic observations will be discussed.

COMP 67 [818976]: Computation of zero-field splitting in triplet heteroarylnitrenes

Zdenek Havlas¹, Mojmir Kyvala¹, and Josef Michl². (1) Institute of Organic Chemistry and Biochemistry, Academy of Sciences of Czech Republic, 166 10 Prague 6, Czech Republic, (2) Department of Chemistry, University of Colorado, Boulder, CO 80309, Fax: 303-492-0799, michl@eefus.colorado.edu

We report the zero-field splitting parameters D and E of a series of heteroaromatic triplet arylnitrenes Ar-N computed at the CASSCF(14,14)/cc-pVDZ to CASSCF(14,11)/cc-pVDZ levels at B3LYP/cc-pVTZ geometries, using the full Breit-Pauli Hamiltonian (Ar = phenyl, 2-pyridyl, 3-pyridyl, 2-pyrimidyl, 2-pyrazinyl, 3-pyridazinyl, and 1,3,5-triazinyl). A comparison of spin-spin dipolar and spin-orbit coupling contributions shows that the latter increase the D value by about 10% and have an irregular effect on the small E value. Experimental values had been measured on nitrenes prepared by irradiation of matrix isolated aryl azides [1-3] and are 10-15% lower than those presently calculated. The triplet species produced from 3-pyridazyl azide is exceptional in that its D value does not agree with that calculated for 3-pyridazylnitrene at all, and we propose that it has another structure.

[1] Kuzaj, M; L�erssen, C.; Wentrup, C. Angew. Chem. Int. Ed. Engl. 1986, 25, 480. [2] Wasserman, E. Prog. Phys. Org. Chem. 1971, 8, 319. [3] Wentrup, C.; Kvaskoff, D., private communication.

COMP 68 [822592]: Calculation of the dispersion energy between large molecules: Graphene plates and the graphene-water system

Peter Pulay and Alan R. Ford, Department of Chemistry and Biochemistry, University of Arkansas, Fayetteville, AR 72701, Fax: 479-575-4049, pulay@uark.edu

We have recently developed an efficient parallel canonical second-order Moller-Plesset (MP2) program. This program, and related newer developments: the calculation of MP2 forces, and current work on an efficient parallel coupled cluster code, will be discussed, with particular attention to parallel disk storage. We have studied the interaction between graphene plates, and between graphene plates and water. The efficiency of our code made it possible to use large models, e.g. two circumcoronene molecules, C₅₄H₁₈ with fairly large basis sets (about 2000 basis functions) on a modest size cluster. MP2 overestimates the dispersion attraction between aromatic systems. However, spin-component scaled MP2 (SCS-MP2) is in good agreement with high-level results, e.g. CCSD(T), for smaller model systems. The results will be discussed in the light of the hydrophobic effect and water depletion between two closely spaced hydrophobic surfaces, and models of interaction between graphitic surfaces.

COMP 69 [818389]: A complete basis set model chemistry for excited states

George A. Petersson, Hall-Atwater Laboratories of Chemistry, Wesleyan University, Middletown, CT 06459-0180, Fax: 860-685-2211, george@dali.wesleyan.edu

Extrapolation schemes for calculations employing a single reference configuration are now used routinely. We have begun the development of complete one-electron basis set (CBS) extrapolation schemes for multi-configuration methods. We select the full valence complete active space self consistent field (CASSCF) multiconfiguration reference for a CISD calculation of the dynamic correlation energy. Extrapolation requires a well defined sequence of approximations and a model for the convergence of this sequence. The convergence of full valence CASSCF energies to the complete basis set limit is very similar to the basis set convergence of UHF energies. This similarity is exploited with extrapolations employing quadruple-ζ UHF energies to extrapolate double- and triple-ζ CASSCF calculations. In an analogous fashion, well established MP2 correlation energy extrapolations are used to estimate the dynamic error in CASSCF-CISD calculations. The six lowest energy singlet and triplet states of the N₂ molecule provide a well documented set of test cases for this method. Preliminary results for the calculated (and experimental) excitation energies, T_e in eV, from the N₂ X¹Σ_g⁺ ground state to the A³Σ_u⁺, B³Π_g, W³Δ_u, B'³Σ_u⁺, a'¹Σ_u^-, a¹Π_g, and w¹Δ_u excited states are: 6.247 (6.224), 7.406 (7.392), 7.468 (7.415), 8.193 (8.217), 8.494 (8.450), 8.573 (8.590), and 9.005 (8.939) respectively.

COMP 70 [824994]: Quantum photochemistry

Donald G. Truhlar¹, Ahren W. Jasper¹, Shikha Nangia¹, Chaoyuan Zhu¹, Piotr Piecuch², and Michael J. McGuire². (1) Department of Chemistry and Supercomputing Institute, University of Minnesota, 207 Pleasant St. SE, Minneapolis, MN 55455-0431, Fax: 612-626-9390, truhlar@umn.edu, (2) Department of Chemistry, Michigan State University

We will discuss the two steps in the calculation of the dynamics of photochemical processes: (1) obtaining the potential energy surfaces and the terms coupling them; (2) calculation of the dynamics itself. For step 1 we use the fourfold way [H. Nakamura and D. G. Truhlar, Journal of Chemical Physics 118, 6816 (2003)] for the direct calculation of diabatic surfaces and couplings. We will present results for the photodissociation of ammonia as an example of the method. For step 2 we use non-Born-Oppenheimer trajectories calculated by coherent switching with decay of mixing [C. Zhu, S. Nangia, A. W. Jasper, and D. G. Truhlar, Journal of Chemical Physics 121, 7858 (2004)]. We will present tests against accurate quantum dynamics for weakly interacting surfaces, avoided crossings, and conical intersections and discuss the effect of treating coherence and decoherence in different ways. This work was supported in part by the National Science Foundation.

COMP 71 [825418]: Advanced computational methods applied to chemistry

Theresa L. Windus¹, Yuri Alexeev¹, Edoardo Apra¹, Manojkumar Krishnan², Vinod Tipparaju², Bruce J. Palmer², and Jarek Nieplocha². (1) Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, 902 Battelle Boulevard, P.O. Box 999, MSIN: K8-91, Richland, WA 99352, Fax: 509-375-6631, theresa.windus@pnl.gov, (2) Computational Science and Mathematics, Pacific Northwest National Laboratory

John Pople was one of the driving forces behind the development of widely used computer codes to solve important chemical problems. Our work will describe how advanced computational science techniques are continuing to be applied to solve large, challenging chemical issues. In particular, we will discuss the use of the common component architecture and multiple levels of parallelism within NWChem to enable dynamic environments for performing large computations on high performance architectures. Several applications will be discussed to show the flexibility and viability of the software.

COMP 72 [828291]: Practical approaches to condensed phase quantum dynamics based on quantum decoherence

Peter J. Rossky, Institute for Theoretical Chemistry, Dept. of Chemistry & Biochemistry, University of Texas at Austin, 1 University Station A5300, Austin, TX 78712, Fax: 512-471-1624, rossky@mail.utexas.edu

The computationally convenient description of the quantum mechanical dynamics of electronic systems and of light nuclei in condensed phases is an important goal for the accurate atomistic representation of chemical and biochemical processes. In electronic systems, the rate of decay of an excited state is influenced by the coherence among the amplitudes for the initial state and the component decay channels. For a species in a condensed phase, the dynamics of the bath can strongly dissipate this coherence and thus modify the rate of electronic evolution. For nuclear dynamics, the coherence between alternative nuclear paths has a parallel significance, and interaction with the surroundings has a similar dissipating effect. Practical approaches to simulation of these cases, which take advantage of the loss of coherence, will be described. These only require evaluation of the classical molecular dynamics of nuclei. The evolution of electronic excited states and the description of the dynamics of light nuclei, including the dissipative role of the environment, will be illustrated in practical cases, including electronic excited state relaxation in solution and in isolated large molecules, and motion in neat liquid para-H2 and He(4).

COMP 73 [832798]: Linearized path integral approach to calculating non-adiabatic time correlation functions

David F Coker, Department of Chemistry 590 Commonwealth Ave., Boston University, 590 Commonwealth Avenue, Boston, MA 02215, Fax: 617-353-6466, coker@bu.edu, and Sara Bonella, Department of Chemistry, Boston University

In statistical mechanics, time correlation functions are central quantities which bridge the microscopic dynamics and fluctuations of a given system to macroscopic, phenomenological quantities, such as transport coefficient or relaxation times. While relatively standard numerical methods provide a viable tool for their evaluation for classical systems, full quantum mechanical calculations of time correlations functions are currently out of the realm of affordable computations. Consequently, many approximate techniques have been developed to tackle this problem. In this presentation we outline a new mixed quantum-classical approach, belonging to the family of the so-called linearization methods, which addresses the evaluation of time correlation functions of nuclear or electronic operators evolving in the presence of non-adiabatic effects. We begin by rewriting the function in a basis set defined as the tensor product of nuclear positions and diabatic electronic states. The approach involves linearizing in the difference between forward and backward nuclear paths while keeping keeping all orders in the electronic occupation paths described using the Mapping Hamiltonian formulation. Results are present for excited state reaction dynamics in various model condensed phase systems.

COMP 74 [828830]: Electrochemistry in a very small cell: A computational approach

Michiel Sprik¹, Jochen Blumberger¹, and Yoshitaka Tateyama². (1) Department of Chemistry, University of Cambridge, Lensfield Road, CB2 1EW, Cambridge, United Kingdom, Fax: +44-1223-336362, ms284@cam.ac.uk, (2) National Institute for Materials Science

Electrochemists control the thermodynamic driving force of redox reactions in an electrochemical cell by applying a voltage to the electrodes. In computations the same effect can be achieved by shifting the potential energy surfaces of different oxidations states with respect to each other (omitting any physical electrode). We have implemented this method using density functional theory based ab initio molecular dynamics simulation ("Car-Parrinello")[1]. In this talk we summarize the latest results we have obtained applying this method to a number of small redox active aqueous solutes (mainly transition metal complexes). We compare our computations to experiment (redox potentials) as well as theory (Marcus Theory of electron transfer.)

[1] J. Blumberger, L. Bernasconi, I. Tavernelli, R. Vuilleumier, M. Sprik, J. Am. Chem. Soc. 126, 3928 (2004)

COMP 75 [807803]: A coarse grain simulation methodology for structured solutions

John C. Shelley, Schrodinger Inc, 1500 SW First AVenue, Suite 1180, Portland, OR 97006, Fax: 503-299-4532, jshelley@schrodinger.com

A methodology for developing coarse grain models for simulating specific structured solutions will be presented. In this approach coarse grain sites, corresponding to roughly 10 atoms, are selected so as to mimic the overall framework of the molecules of interest. The intramolecular and intermolecular interactions are parameterized so as to reproduce data from experiment and aspects of intra site distributions obtained from atomistic simulations. The resulting models have reproduced key phenomena and have proved useful in predictive studies. While this is encouraging the model will be examined critically to highlight its strengths and weaknesses with an eye towards prospects for future enhancements.

COMP 76 [829452]: The performance of metadynamics in flexible docking

Francesco L. Gervasio, Physical Chemistry, ETH Zuerich, Research group Prof. Parrinello, Usi Campus, CH-6900 Lugano, Switzerland, Fax: +41 (91) 9138817, fgervasi@phys.chem.ethz.ch, Allesandro Laio, Research Group Prof. Parrinello, ETH Zurich, and Michele Parrinello, Physical Chemistry ETH (Zurich)

We apply our recently developed metadynamics method to the docking of ligands on flexible receptors in water solution. This method mimics the real dynamics of a ligand exiting or entering an enzyme and in so doing reconstructs the free energy surface. We apply it to four classical docking cases: beta-trypsin/benzamidine, beta-trypsin/ chloro-benzamidine, immunoglobulin McPC-603/phosphocoline and CDK2/staurosporine. In every case studied the method is able to predict the docked geometry and the free energy of docking. Its added value with respect to many other available methods is that it reconstructs the complete free energy surface including all the relevant minima and the barriers between them.

COMP 77 [819284]: Binding MOAD (Mother of All Databases)

Heather A. Carlson¹, Mark L. Benson², Richard D. Smith³, and Liegi Hu¹. (1) Department of Medicinal Chemistry, University of Michigan, Ann Arbor, 428 Church Street, Ann Arbor, MI 48109, carlsonh@umich.edu, (2) Bioinformatics Program, University of Michigan, Ann Arbor, (3) Biophysics Research Division, University of Michigan, Ann Arbor

There is a wealth of information about protein-ligand interactions contained in the PDB. We present a comprehensive database of those structures: Binding MOAD (Mother of All Databases). The methods involved in currating the dataset will be covered. Binding MOAD contains 5359 valid protein-ligand complexes. There are 2660 unique ligands and 2090 unique protein families. Over 5000 crystallography papers were searched for binding data, and 1378 (26%) of the complexes are augmented by binding affinity information. We have mined the dataset for information about the biophysics of molecular recognition. The binding sites have been analyzed for amino acid content, geometry, atomic contacts, ligand volume, binding site volume, and degree of solvent exposure. These are then compared to the large number of Ki, Kd, or IC50 values to correlate the size and makeup of the binding site with the binding affinity.

COMP 78 [833580]: SAR, binding mode and radioprotective effects of a novel class of Checkpoint2 kinase inhibitors

Frank U. Axe¹, Kristen L. Arienti², Anders Brunmark³, Kelly McClure⁴, Alice Lee⁵, Jon Blevitt³, Danielle Neff³, Liming Huang⁴, Shelby Crawford³, Chennagiri R. Pandit³, Lars Karlsson³, and J. Guy Breitenbucher⁶. (1) Department of Chemistry, Johnson and Johnson Pharmaceutical Research and Development, 3210 Merryfield Row, San Diego, CA 92121, (2) Department of Chemistry, Johnson and Johnson Pharmaceutical Research and Development, LLC, (3) (4) Department of Chemistry, Johnson & Johnson Pharmaceutical Research & Development, L.L.C, (5) Department of Chemistry, University of California at Berkeley, (6) Pharmaceutical Research Institute, RW Johnson

Checkpoint2 (Chk2) kinase is a potentially important target in radiation treatment of cancer patients. A novel series of 2-arylbenzimidazoles were found to inhibit this target selectively. Molecular docking was used to predict the binding mode of this novel kinase chemotype in the ATP pocket of Chk2. This binding model was used to rationalize the SAR and to develop a refinement strategy to improve potency and optimize other medicinal properties. These compounds were found to be ATP competitive and provide dose dependent protection of human CD4+ and CD8+ T-cells from apotosis due to ioninzing radiation.

COMP 79 [832944]: Understanding protein-ligand interactions in the field of kinases

Isabelle Morize, Dorothea Kominos, and Robert Pearlstein, Molecular Modeling, Sanofi-Aventis, Route 202-206, PO Box 6800, JR1-203A, Bridgewater, NJ 08807, Fax: 908-231-3577, isabelle.morize@aventis.com

Kinases are a challenge for drug design, due to multiple factors: one of which being the flexibility exhibited by these proteins. Understanding protein-ligand interactions is one of the primary steps in designing new compounds with improved affinity, along with the right properties to make them drugs. Working on multiple targets from the same family, such as kinases, and getting access to structural information, has been an advantage for modelers, since it has afforded them a knowledge-driven approach to exploit common patterns, highlight differences, and rapidly apply this knowledge via proprietary structures displaying drug-like properties with regard to affinity, selectivity, toxicity, absorption, distribution, metabolism and excretion. Navigation into the biological and chemical spaces, in order to identify the best intersection (i.e. best targets/best compounds) is one of our primary objectives, and we will present the lessons we learned during this voyage.

COMP 80 [834602]: Characterization of protein-ligand interaction sites using computational solvent mapping

Sandor Vajda¹, Michael Silberstein², and Karl Clodfelter². (1) Department of Biomedical Engineering, Boston University, 44 Cummington St, Boston, MA 02215, Fax: 617-353-6766, vajda@bu.edu, (2) Program in Bioinformatics, Boston University

Computational solvent mapping moves molecular probes - small organic molecules containing various functional groups – around the protein surface, finds favorable positions using empirical free energy functions, clusters the conformations, and ranks the clusters on the basis of the average free energy. The mapping procedure reproduces the available experimental solvent mapping results, eliminating the problem of spurious local minima associated with previous computational methods. The most important result is that using at least six different probes, the consensus site at which most probes overlap is always a major subsite of the functional site. In addition, the amino acid residues that interact with the probes also bind the specific ligands of the protein, and thus the method provides detailed and reliable information on the functional sites. We apply the approach to cytochrome P450s and peroxisome proliferator activated receptors (PPARs), and show that it yields substantial biological insight and facilitates drug design.

COMP 81 [834204]: End point free energy calculations: What it takes for success with FK506 Binding Protein

Jessica M. J. Swanson, Department of Chemistry and Biochemistry, University of California San Diego, 9500 Gilman Dr, La Jolla, CA 92037-0365, Fax: 858-534-7042, jswanson@mccammon.ucsd.edu, and J. Andrew McCammon, Howard Hughes Medical Institute, Department of Chemistry and Biochemistry, University of California at San Diego

End-point free energy calculations have received increasing attention over the last decade. These methods combine gas phase energies from explicit solvent simulations with continuum solvation energies to evaluate the free energy of the bound and free states of a binding reaction. This work explores the importance of using compatible implicit and explicit solvent models. We report the effects of using different continuum parameter sets including a set of radii that have been recently optimized for Poisson-Boltzmann calculations with the AMBER (parm99) partial charges. These radii were benchmarked against explicit solvent electrostatic solvation energies to provide the optimal explicit and implicit energetic agreement. Different radii were reported for abrupt and cubic spline smoothed surface definitions. We report the effect of these surface definitions on the measured free energies as well as their relative computational demand.

COMP 82 [820166]: Dynamics of buried water in proteins

Chandra S Verma, Computational Biology, Bioinformatics Institute, 30 Biopolis Way, #07-01 Matrix, Singapore 138671, Singapore, Fax: 65-6478-9077, chandra@bii.a-star.edu.sg

Using water as a paradigm for ligand binding we find that the the vibrational spectrum of the protein, particularly the low freqeuncy modes, undergoes a red shift upon ligand binding; this leads to entropic stabilization which is in contrast to the classical notion that ligand binding induces a tightening of the protein. Further, the pathways of entry of this liagdn into and out of the protein are characterized by a rugged energy landscape that requries a departure in analysis from the standard two-state paradigm underlying the Arrhenius scenario. This methodology is then extended to resolving a long standing disagreement between NMR and crystallographic data on whether waters can exist in hydrophobic cavities in proteins, by finding in the affirmative.

COMP 83 [833478]: A new approach to efficiently explore the free energy surface, trace the reaction pathways and converge the free energy barriers

Bernd Ensing, Center for Molecular Modeling, Department of Chemistry, University of Pennsylvania, 231 S. 34th Street, Philadelphia, PA 19104, Fax: 215-573-6233, ensing@cmm.upenn.edu, Alessandro Laio, Research Group Prof. Parrinello, ETH Zurich, Michele Parrinello, Computational Science, Department of Chemistry and Applied Biosciences, ETH Zurich, and Michael L. Klein, Department of Chemistry, University of Pennsylvania

The recently introduced hills method is a powerful tool to compute the multi-dimensional free energy surface of intrinsically concerted reactions. We have extended this method by focusing our attention on localizing the lowest free energy path that connects the stable reactant and product states. This path represents the most probable reaction mechanism, similar to the zero temperature IRC, but also includes finite temperature effects. The transformation of the multi-dimensional problem to a one-dimensional reaction coordinate allows for accurate convergence of the free energy profile along the lowest free energy path using standard free energy methods. The power of the method is exemplified by a number of interesting chemical and biophysical problems, among which the concerted enzyme catalyzed phosphate transfer reaction, the anti-microbial peptide nanotube insertion into a lipid bilayer and the exploration of the free energy landscape of the reactions between F- and CH3CH2F showing the E2 and SN2 reaction channels simultaneously.

COMP 84 [833231]: Application of digital filters to enhance conformational change in protein systems

Adrian P. Wiley and Jonathan W Essex, School of Chemistry, University of Southampton, Highfield, Southampton SO17 1BJ, United Kingdom, Fax: +44 (0)23 8059 3781, jwe1@soton.ac.uk
Motions in biological molecules can occur on timescales beyond those accessible to molecular dynamics simulations, and algorithms that enhance conformational sampling are therefore of considerable interest. Reversible Digitally Filtered Molecular Dynamics (RDFMD) is able to enhance sampling by amplifying the low frequency motions as a simulation evolves. In this presentation, methods to optimise the performance of RDFMD are described, and applications of this approach to enhancing the conformational motion in E. coli dihydrofolate reductase, T4 lysozyme, and HIV-1 protease, are described.

COMP 85 [827552]: Atomically detailed simulations of conformational transitions in DNA Polymerases: Implications for DNA synthesis fidelity mechanisms

Karunesh Arora, Department of Chemistry, New York University, 31 Washington Place, Silver Center for Arts and Science, New York, NY 10003, Fax: 212-995-4152, ka357@nyu.edu, and Tamar Schlick, Department of Chemistry, Courant Institute of Mathematical Sciences, and the Howard Hughes Medical Institute, New York University

DNA polymerases play a key role in maintaining genome integrity by repairing damaged DNA bases. If the damaged DNA is replicated unrepaired it may lead to cancer and premature aging. The fidelity of polymerases depends on their ability to incorporate correct rather than incorrect nucleotides complementary to the template DNA; such fidelities lie in the range from 1 to nearly 10⁶ errors per million nucleotides incorporated. Based on kinetic and structural data, it is likely that high fidelity enzymes (like pol beta) undergo conformational changes prior to nucleotide incorporation such as to tailor-fit correct base-pairs rather than incorrect units, while low fidelity enzymes (like Dpo4) discriminate correct from incorrect base-pairs very poorly due to a much more open active site. To this end, we have developed and applied a combination of molecular dynamics and novel long-time dynamics simulation methodologies to determine conformational transition pathways of Pol beta and Dpo4 in the presence of correct and incorrect incoming substrates in the active site. These simulations have helped determine the order of events, slow reaction coordinates, and key transition state regions in the presence of matched and mismatched base-pairs. We observe that in the presence of mismatched base-pairs, the N-subdomain of pol beta which is believed to be required for catalytic cycling, remains in the "open" conformation, rather than the "closed" conformation observed in the presence of correct base-pair. In contrast to pol beta, in Dpo4 we see the sliding of the DNA template/primer strands in response to the incoming substrate in the active site. Protein subdomains undergo conformational changes that are smaller in magnitude and distinct in identity from pol beta.

COMP 86 [828826]: Hybrid molecular dynamics-quantum mechanics simulations of solvation dynamics

Matthew C. Zwier, Christopher M. Meeusen, Justin M. Shorb, and Brent P. Krueger, Department of Chemistry, Hope College, 35 East 12th St., Holland, MI 49423, kruegerb@hope.edu

Solvation dynamics play a large role in chemical processes, whether a traditional small molecule in solution or a cofactor surrounded by protein. Recently, a hybrid molecular dynamics (MD) and quantum mechanical (QM) method (MD/QM) has been developed by Mercer, Gould, and Klug that may be both more applicable to complex systems and more robust than purely classical methods. While it has been used to simulate the (linear) absorption spectra of a number of complex biological systems, to date only limited basic evaluation of the method has been completed. We present the results of MD/QM calculations on an organic dye (oxazine 4) in methanol, compare them to both linear and nonlinear optical spectra, and evaluate a number of critical parameters such as simulation length, QM sampling rate, and QM model chemistry. In particular, a computationally inexpensive semi-empirical QM method gives encouraging results in contrast to previous studies.

COMP 87 [818470]: Long dynamics simulations of proteins using atomistic force fields and a continuum representation of solvent effects

Xianfeng Li¹, Sergio A. Hassan², and Ernest L. Mehler¹. (1) Physiology and Biophysics, Weill Medical College of Cornell University, 1300 York Avenue, New York, NY 10021, elm2020@med.cornell.edu, (2) National Institutes of Health, DHHS

Long molecular dynamics (MD) simulations were carried out on the B1 immunoglobulin-binding domain of streptococcal protein G (ProtG) and bovine pancreatic trypsin inhibitor (BPTI) using an atomistic force field and a continuum representation of solvent effects. To mimic frictional and random collision effects, Langevin dynamics (LD) were used. The main goal of the calculations was to explore the stability of tens-of-nanosecond trajectories using a continuum model, and to analyze, in detail, structural and dynamical properties. Conformational fluctuations, order parameters, cross correlation matrices, residue solvent accessibilities, pKa values of titratable groups, and hydrogen-bonding (HB) patterns were calculated from all the trajectories and compared with available experimental data. The simulations comprised 30ns and 40ns for BPTI and ProtG, respectively. For comparison explicit water (EW/MD) of 4 ns and 3 ns, respectively, were also carried out. Two continuum simulations were performed on each protein using CHARMM: one with the all-atom PAR22 representation of the protein forcefield (referred to as PAR22/LD) and the other with the recently developed CMAP potential (CMAP/LD). The continuum model is based on the screened Coulomb potential (SCP) reported earlier, the SCP-based implicit solvent model (SCP-ISM). For ProtG both the PAR22/LD and CMAP/LD 40ns-trajectories were stable, but for BPTI only the CMAP/LD trajectory was stable for the entire 30-ns simulation. The source of the instability of the PAR22/LD simulation of BPTI was explored by an analysis of the backbone torsion angles.

COMP 88 [830533]: Constant pH molecular dynamics in generalized Born implicit solvent

John T. Mongan, Bioinformatics, Medical Scientist Training Program, NSF Ctr. Theor. Biol. Physics, University of California San Diego, 9500 Gilman Dr. #0365, La Jolla, CA 92093-0365, Fax: 858-534-4974, jmongan@mccammon.ucsd.edu, David A. Case, Dept. of Molecular Biology, Scripps Research Institute, and J Andrew McCammon, Howard Hughes Med. Inst., NSF Ctr. Theor. Biol. Physics, Dept. Chem. and Biochem., and Dept. Pharmacol, University of California, San Diego

Traditional molecular dynamics (MD) methods employ fixed protonation states and so cann

The Computers in Chemistry Division Spring 2005 ACS Meeting Abstracts

COMP Programming

229th ACS National Meeting San Diego, CA March 13 - 17, 2005 W. D. Cornell, Program Chair

The Computers in Chemistry Division
Spring 2005 ACS Meeting Abstracts

229th ACS National Meeting
San Diego, CA
March 13 - 17, 2005
W. D. Cornell, Program Chair