The generalized hybrid orbital method for combined quantum mechanical/molecular mechanical calculations: formulation and tests of the analytical derivatives

 Hybrid quantum mechanical (QM) and molecular mechanical (MM) potentials are becoming increasingly important for studying condensed-phase systems but one of the outstanding problems in the field has been how to treat covalent bonds between atoms of the QM and MM regions. Recently, we presented a generalized hybrid orbital (GHO) method that was designed to tackle this problem for hybrid potentials using semiempirical QM methods [Gao et al. (1998) J Phys Chem A 102: 4714–4721]. We tested the method on some small molecules and showed that it performed well when compared to the purely QM or MM potentials. In this article, we describe the formalism for the determination of the GHO energy derivatives and then present the results of more tests aimed at validating the model. These tests, involving the calculation of the proton affinities of some model compounds and a molecular dynamics simulation of a protein, indicate that the GHO method will prove useful for the application of hybrid potentials to solution-phase macromolecular systems. Received: 4 October 1999 / Accepted: 18 December 1999 / Published online: 5 June 2000     

Implementation of the IMOMM methodology for performing combined QM/MM molecular dynamics simulations and frequency calculations

A technique for implementing the integrated molecular orbital and molecular mechanics (IMOMM) methodology developed by Maseras and Morokuma that is used to perform combined quantum mechanics/molecular mechanics (QM/MM) molecular dynamics simulations, frequency calculations and simulations of macromolecules including explicit solvent is presented. Although the IMOMM methodology is generalized to any coordinate system, the implementation first described by Maseras and Morokuma requires that the QM and MM gradients be transformed into internal coordinates before they are added together. This coordinate transformation can be cumbersome for macromolecular systems and can become ill-defined during the course of a molecular dynamics simulation. We describe an implementation of the IMOMM method in which the QM and MM gradients are combined in the cartesian coordinate system, thereby avoiding potential problems associated with using the internal coordinate system. The implementation can be used to perform combined QM/MM molecular dynamics simulations and frequency calculations within the IMOMM framework. Finally, we have examined the applicability of thermochemical data derived from IMOMM framework. Finally, we have examined the applicability of thermochemical data derived from IMOMM frequency calculations. Received: 11 May 1998 / Accepted: 14 August 1998 / Published online: 16 November 1998     

Second-order Møller–Plesset perturbation theory with terms linear in the interelectronic coordinates and exact evaluation of three-electron integrals

 The second-order correlation energy of M?ller–Plesset perturbation theory is computed for the neon atom using a wave function that depends explicitly on the interelectronic coordinates (MP2-R12). The resolution-of-identity (RI) approximation, which is invoked in the standard formulation of MP2-R12 theory, is largely avoided by rigorously computing the necessary three-electron integrals. The basis-set limit for the second-order correlation energy is reached to within 0.1 mE _h. A comparison with the conventional RI-based MP2-R12 method shows that only three-electron integrals over s and p orbitals need to be computed exactly, indicating that the RI approximation can be safely used for integrals involving orbitals of higher angular momentum. Received: 9 May 2001 / Accepted: 31 October 2001 / Published online: 9 January 2002     

Molecular dynamics, density functional and second-order Møller–Plesset theory study of the structure and conformation of acetylcholine in vacuo and in solution

 Molecular mechanics minimizations based on the CVFF force field and molecular dynamics simulation for a time of 2.5 ns were performed to examine the conformational behaviour and the molecular motion of acetylcholine in vacuo and in aqueous solution. Five low-lying conformations, namely the TT, TG, GG, G*G and GT, were obtained from molecular mechanics computations with the GT structure as the absolute minimum. Molecular dynamics trajectories in vacuo and in water show that only four (GT, GG, G*G and TG) and three (TG, TT and GT) conformations are present in the simulation time, respectively. Density functional B3LYP and second-order M?ller–Plesset (MP2) methods were then used to study all the five lowest-lying conformers of acetylcholine neurotransmitter in vacuo and in water by the polarizable continuum model approach. The B3LYP and MP2 computations show that in the gas phase all minima lie in a narrow range of energy with the G*G conformer as the most stable one. The relative minima GG, GT, TG and TT are located at 1.1 (3.3), 1.8 (4.2), 2.1 (4.5) and 4.3 (7.3) kcal/mol above the absolute one at the B3LYP (MP2) level. The preferred conformation in water is the TG. Solvation reduces the relative energy differences between the five minima in both computations. Received: 4 April 2001 / Accepted: 5 July 2001 / Published online: 30 October 2001     

Hydration of platinum(II) complexes: a molecular mechanics study using atom-based force-field parameters

 This work is related to the interaction of water with two platinum(II) complexes, [Pt(NH₃)₄]²⁺ (denoted 1) and trans-[Pt(OH)₂(NH₃)₂] (denoted 2). We have considered two approaches of a water molecule to complexes 1 and 2 along the z-axis normal to the platinum(II) coordination plane: approach I, with the water oxygen oriented towards Pt, and approach II, with one water hydrogen directed towards Pt. Calculations have been performed within a molecular mechanics method based upon the interaction potentials proposed earlier by Claverie et al. and subsequently adjusted to results obtained with symmetry – adapted perturbational theory as well as with supermolecule (up to second-order M?ller–Plesset, MP2) methods. We discuss some possible simplifications of the potentials mentioned. The results relative to the hydration of Pt complexes 1 and 2 following approach I or II are discussed and compared to recent (MP2) ab initio energy–distance curves that we have recently determined. The MP2 calculations have shown that besides exchange–repulsion contributions, which are very similar in all hydrated complexes, approach I is mainly governed by electrostatics, whereas for approach II both electrostatic and dispersion contributions are important. Received: 16 September 1999 / Accepted: 3 February 2000 / Published online: 5 June 2000     

Structure and stabilities of (HA1NH)n (n = 2–4)

Ab initio molecular electronic structure methods have been used to study the relative stability of the planar inorganic ring (HAlNH)_n (n = 2–4) during homodesmotic and monomer polymerization reactions. Optimized geometries, frequencies and energies through restricted Hartree-Fock/6-31G* are reported, and energies at the self-consistent field optimized geometries including M?ller-Plesset fourth perturbation theory with single, double and quadruple excitation (MP4SDQ) corrections are also reported for both reactions. Homodesmotic reactions with MP4SDQ −28.5 kcal/mol for (AlN)₂, 1.9 kcal/mol for (AlN)₃ and −0.97 kcal/mol for (AlN)₄. On analysing a π-molecular orbitals diagram, only one, three and three strongly bonding π-molecular orbitals exist for the planar four-, six- and eight-membered AlN rings, respectively. Received: 9 March 1998 / Accepted: 19 March 1998 / Published online: 23 June 1998     

Theoretical study of the potential-energy surface of C2NP

 The B3LYP/6-311G(d) and CCSD(T)/6-311G(2df) (single-point) methods have been used to investigate the singlet potential energy surface of C₂NP, in which seven stationary isomers and seventeen interconversion transition states are involved. At the final CCSD(T)/6-311G(2df)//B3LYP6-311G(d) level with zero-point vibrational energy correction the lowest-lying isomer is a linear NCCP followed by two linear CNCP isomers at 23.9  and CCNP at 65.8 kcal mol⁻¹, respectively. The three isomers are kinetically very stable towards both isomerization and dissociation, and CCNP is even more kinetically stable than CNCP – by 14.3 kcal mol⁻¹ despite its high energy. Further comparative calculations were performed at the QCISD and QCISD(T) levels with the 6-311G(d) and 6-311G(2d) basis sets to obtain more reliable structures and spectroscopy for the three isomers. The calculated bond lengths, rotational constant, and dipole moment for NCCP were in close agreement with the experimentally determined values. Finally, similarities and discrepancies between the potential energy surface of C₂NP and those of the analogous species C₂N₂ and C₂P₂ were compared. The results presented in this paper might be helpful for future identification of the two still unknown yet kinetically very stable isomers CNCP and CCNP, both in the laboratory and in interstellar space. Received: 3 January 2001 / Accepted: 6 June 2001 / Published online: 30 October 2001     

Dual-level direct dynamics calculations of deuterium kinetic isotope effects for the Cl(2P)+C2H6 abstraction reaction

 Kinetic isotope effects, KIEs, for hydrogen abstraction from C₂H₆ and C₂D₆ by chlorine atom have been studied by the dual-level direct dynamics approach. A low-level potential energy surface is obtained with the MNDO-SRP method. High-level structural properties of the reactants, transition state, and products were obtained at the MP2 level with the cc-pVDZ, aug-cc-pVDZ, and the cc-pVTZ basis sets. Using the variational transition state theory with microcanonical optimized multidimensional tunneling, the values of deuterium KIE, at 300 K, range from 2.28 to 3.27, in good agreement with the experimental values (2.69–5.88). Received: 6 June 2001 / Accepted: 12 July 2001 / Published online: 19 November 2001     

Model studies of the chemisorption of hydrogen and oxygen on the Au (1 0 0) surface

 The convergence of chemisorption energy for hydrogen and oxygen on gold clusters is studied. Two theoretical approaches have been employed; wavefunction methods at the self-consistent-field second–order M?ller–Plesset level and density functional theory and the two methods are compared. Relativistic effective core potentials exploited in the former approach were developed in this work. Received: 25 October 1999 / Accepted: 21 February 2001 / Published online: 11 October 2001     

Reparameterization of hybrid functionals based on energy differences of states of different multiplicity

 Low-spin/high-spin energy splittings for Fe(II) transition-metal complexes – particularly in weak ligand fields – cannot be well described by density functional methods. Different density functionals yield results which differ by up to 1 eV in transition-metal complexes with sulfur-rich first coordination spheres. We attribute this failure to the fact that the high-spin state is systematically favoured in Hartree–Fock-type theories, because Fermi correlation is included in the exact exchange, while Coulomb correlation is not. We thus expect that the admixture of exact exchange to a given density functional will heavily influence the energy splitting between states of different multiplicity. We demonstrate that the energy splitting depends linearly on the coefficient of exact exchange admixture. This remarkable result is found for all the Fe(II)–S complexes studied. From this observation we conclude in connection with experimental results that Becke's 20% admixture should be reduced to about 15% if meaningful energetics are sought for transition-metal compounds. We rationalize that this reduction by 5% will not affect the quality of the hybrid functional since we arrive at a slightly modified functional, which lies between the pure density functional and the hybrid density functional, which both give good results for “standard” systems. Received: 13 July 2001 / Accepted: 31 August 2001 /  Published online: 16 November 2001     

Examining ligand-protein interactions with binding-energy landscapes

Binding-energy landscapes are used to investigate the thermodynamics of molecular recognition for the pteridine ring, a recognition anchor in binding with dihydrofolate reductase, and two molecules with the same shape but different heteroatom substitutions. The relative importance of hydrogen bonding and hydrophobic interactions in this system is analyzed by comparing these three different decorations of the pteridine scaffold. Received: 5 May 1998 / Accepted: 3 September 1998 / Published online: 17 December 1998     

Comparison of solvent reaction field representations

 Alternative ways are examined for representing a reaction field to treat the important effects of long-range electrostatic interaction with a solvent in electronic structure calculations on the properties of a solute. Several extant boundary element methods for approximate representation of the solvent reaction field in terms of surface charge distributions are considered, and analogous new methods for approximate representation in terms of surface dipole distributions are introduced. Illustrative computational results are presented on representative small neutral and ionic solutes to evaluate the relative accuracy of various methods. Received: 2 July 2001 / Accepted: 10 September 2001 / Published online: 19 December 2001     

Molecular modeling study of an abasic DNA undecamer duplex: d(GCGTGOGTGCG) · d(CGCACTCACGC)

Molecular mechanics calculations were performed with the JUMNA program on d(GCGTGOGTGCG) · d(CGCACTCACGC) where “O” is a modified abasic site: 3-hydroxy-2-(hydroxymethyl)tetrahydrofuran. From energy minimizations, for intrahelical or extrahelical positions of the unpaired thymine, various structures with different curvatures were obtained. Dynamical properties of this abasic sequence were also investigated through the controlled studies of DNA bending. Poisson-Boltzmann calculations were used to mimic the electrostatic effect of solvent on this sequence. The lowest energy structures show an acceptable agreement with experimental data. Received: 1 June 1998 / Accepted: 17 September 1998 / Published online: 10 December 1998     

Counterpoise-corrected potential energy surfaces of simple H-bonded systems

Geometries and stabilization energies of various simple H-bonded complexes (water dimer, hydrogen fluoride dimer, formamide dimer, formic acid dimer) have been determined by a gradient optimization that eliminates the basis set superposition error (BSSE) by the counterpoise (CP) method in each gradient cycle as well as by the standard gradient optimization. Both optimization methods lead to different potential energy surfaces (PES). The difference depends on the theoretical level used and is larger if correlation energy is considered. Intermolecular distances from the CP-corrected PES are consistently longer, and this difference might be significant (∼0.1 ?); also angular characteristics determined from both surfaces differ significantly. Different geometries were obtained even when passing to larger basis sets (aug-cc-pVDZ). The standard optimization procedure can result in a completely wrong structure. For example, the “quasi-linear” structure of the (HF)₂ (global minimum) does not exist at the standard MP2/ 6-31G** PES (where only cyclic structure was detected) and is found only at the CP-corrected PES. Stabilization energies obtained from the CP-corrected PES are always larger than these from the standard PES where the BSSE is added only a posteriori for the final optimized structure; both energies converge only when passing to a larger basis set (aug-cc-pVDZ). Received: 11 March 1998 / Accepted: 19 June 1998 / Published online: 4 September 1998 RID=" ID=" <E6>Acknowledgements.</E6> The project was supported by the Grant Agency of the Czech Republic (Grant No. 203/98/1166). RID=" ID=" <E5>Correspondence to</E5>: P. Hobza     

Dynamics of the transmembrane domain of the ErbB-2 receptor

The structure and dynamics of the ErbB-2 transmembrane domain have been examined using molecular dynamics techniques both in vacuum and within an explicit hydrated L-α-dilauroyl-phosphatidyl-ethanolamine environment. In-vacuum simulations show that a highly cooperative structural transition occurs frequently within the α-helical transmembrane domain which converts to local π-helices. We show that the α-helix alteration does not depend upon the force field or initial side-chain conformations but is intimately related to the sequence. The membrane-like environment does not prevent the structural transition in the helix but slows down the peptide dynamics indicating that the appearance of a π-bulge is not an artifact of the vacuum approximation. The consequences of π-helix formation could be very huge for the ErbB-2 receptor which is involved in numerous human cancers and also for other membrane proteins wherein similar local structures are also observed experimentally. Received: 9 May 1998 / Accepted: 3 September 1998 / Published online: 17 December 1998     

Determination of a set of parameters for the molecular modelling of phosphorothioate DNA

Phosphorothioate DNAs, have emerged as a new class of potent drugs. They are obtained by the replacement of the anionic oxygens of the phophodiester backbone by sulphur. A set of parameters has been developed for the FLEX force field implemented in JUMNA 10.0 to take into account the influence of sulphur on the structure of the DNA double helix. The consistency of our parameters was tested by modelling a phosphorothioate oligomer namely d(GC)8. d(GC)8. Results, obtained on both R-p_S and S-p_S diastereoisomers, were compared to the phosphodiester counterpart and are in agreement with available experimental data. Thus, our set of parameters seems suitable for further molecular modelling of other phosphorothioate oligomers. Received: 15 July 1998 / Accepted: 8 September 1998 / Published online: 10 December 1998     

The unrestricted natural orbital-restricted active space method: methodology and implementation

The full configuration interaction method in the space of fractionally occupied unrestricted natural orbitals (UNO-CAS method) is extended to excited states as well as to strongly correlated and reactive systems with large active spaces. This is accomplished by␣using restricted active space (RAS) wave functions introduced by Olsen et al. [(1988) J Chem Phys 89: 2185] and using the UNOs without the expensive orbital optimization step. In RAS, the space of active orbitals is subdivided into three groups: a group with essentially doubly occupied orbitals (RAS1), the usual CAS space (RAS2), and a space with weakly occupied active orbitals (RAS3). We select these spaces on the basis of the occupation numbers of the UNOs. All possible electron distributions are allowed in the usual CAS space, but the number of vacancies is limited in RAS1 and the number of electrons is limited in RAS3. We discuss an efficient algorithm for generating a RAS wave function. This is based on the Handy-Knowles determinantal expansion with an addressing scheme adopted for the restricted expansion. Results for both ground and excited states of azulene and free base porphyrin are presented. Received: 16 July 1998 / Accepted: 7 August 1998 / Published online: 19 October 1998     

Direct relativistic MP2: properties of ground state CuF, AgF and AuF

A computer program for the calculation of the MP2 energy correction for a Kramers-restricted Dirac-Hartree-Fock four component wave-function is presented. In the spirit of the integral-driven direct SCF scheme the algorithm has been developed as direct MP2, calculating integrals as they are needed and avoiding the integral storage bottle-neck of conventional MP2. Relativistic MP2 is applied to ground state (¹Σ⁺) CuF, AgF and AuF. Received: 15 December 1996 / Accepted: 2 April 1997     

Ab initio study of rate constants of the reaction: HCN + OH → CN + H2O

A method applying ab initio direct dynamics has been utilized in studying the hydrogen abstraction reaction HCN + OH → CN + H₂O. The geometries of the reactants, products, and the transition state have been optimized at the QCISD/6-311G(d, p) level. Single-point energies were further evaluated at the QCISD(T)/6-311+G(2df, 2p)//QCISD/6-311G(d, p) level. The barrier heights for the forward and reverse reactions were predicted to be 15.95 and 7.51 kcal mol⁻¹ at the QCISD(T)/6-311 + G(2df, 2p)//QCISD/6-311G(d, p) level, respectively. The reaction rate constants were calculated in the temperature range from 298 to 4,000 K using the canonical variational transition-state theory with a small-curvature tunneling correction. The results of the calculation show that the theoretical rate constants are in good agreement with experimental data over the measured temperature range of 400–2,600 K. Received: 18 August 2002 / Accepted: 30 August 2002 / Published online: 20 November 2002 Acknowledgements. Our thanks are due to D.G. Truhlar for providing the POLYRATE 8.2 program. This work was supported by the National Science Foundation of China. We also thank D.C. Fang and Y. M. Xie for their valuable help, and P.R. Yan for reading our paper. Correspondence to: Q. S. Li e-mail: qsli@mh.bit.edu.cn