A well-tempered density functional theory of electrons in molecules

This Invited Article reports extensions of a recently developed approach to density functional theory with correct long-range behavior (R. Baer and D. Neuhauser, Phys. Rev. Lett., 2005, 94, 043002). The central quantities are a splitting functional gamma[n] and a complementary exchange-correlation functional E[n]. We give a practical method for determining the value of gamma in molecules, assuming an approximation for E is given. The resulting theory shows good ability to reproduce the ionization potentials for various molecules. However it is not of sufficient accuracy for forming a satisfactory framework for studying molecular properties. A somewhat different approach is then adopted, which depends on a density-independent gamma and an additional parameter w eliminating part of the local exchange functional. The values of these two parameters are obtained by best-fitting to experimental atomization energies and bond lengths of the molecules in the G2(1) database. The optimized values are gamma = 0.5 a and w = 0.1. We then examine the performance of this slightly semi-empirical functional for a variety of molecular properties, comparing to related works and experiment. We show that this approach can be used for describing in a satisfactory manner a broad range of molecular properties, be they static or dynamic. Most satisfactory is the ability to describe valence, Rydberg and inter-molecular charge-transfer excitations.     

Amine‐Directed Hydrogen‐Bonded Two‐Dimensional Supramolecular Structures

Utilizing pure amine hydrogen bonding is a novel approach for constructing two‐dimensional (2D) networks. Further, such systems are capable of undergoing structural modifications due to changes in pH. In this study, we designed a 2D network of triaminobenzene (TAB) molecules that by varying the pH from neutral to acidic, form either ordered or disordered structures on Au(111) surface as revealed in scanning tunneling microscopy images. In near‐neutral solution (pH ≈5.5), protonation of TAB generates charged species capable of forming H‐bonds between amine groups of neighboring molecules resulting in the formation of a 2D supramolecular structure on the electrified surface. At lower pH, due to the protonation of the amine groups, intermolecular hydrogen bonding is no longer possible and no ordered structure is observed on the surface. This opens the possibility to employ pH as a chemical trigger to induce a phase transition in the 2D molecular network of triaminobenzene molecules.     

Simple pulse sequences for selective multiple-quantum excitation

A sixteen pulse sequence which excites only four-quantum transitions in multiple-quantum NMR is shown to give good selectivity on two molecules, benzene and n-butyl bromide. The behavior as pulse sequence parameters are varied is shown to be consistent with coherent averaging theory. Advantages of this new sequence include much lower duty cycles and shorter total duration than can be achieved with previously used selective sequences. Limitations of this approach and extensions to high-quantum selectivity are discussed.     

The pharmacophore kernel for virtual screening with support vector machines

We introduce a family of positive definite kernels specifically optimized for the manipulation of 3D structures of molecules with kernel methods. The kernels are based on the comparison of the three-point pharmacophores present in the 3D structures of molecules, a set of molecular features known to be particularly relevant for virtual screening applications. We present a computationally demanding exact implementation of these kernels, as well as fast approximations related to the classical fingerprint-based approaches. Experimental results suggest that this new approach is competitive with state-of-the-art algorithms based on the 2D structure of molecules for the detection of inhibitors of several drug targets.     

Generating 3D molecules conditional on receptor binding sites with deep generative models

The goal of structure-based drug discovery is to find small molecules that bind to a given target protein. Deep learning has been used to generate drug-like molecules with certain cheminformatic properties, but has not yet been applied to generating 3D molecules predicted to bind to proteins by sampling the conditional distribution of protein–ligand binding interactions. In this work, we describe for the first time a deep learning system for generating 3D molecular structures conditioned on a receptor binding site. We approach the problem using a conditional variational autoencoder trained on an atomic density grid representation of cross-docked protein–ligand structures. We apply atom fitting and bond inference procedures to construct valid molecular conformations from generated atomic densities. We evaluate the properties of the generated molecules and demonstrate that they change significantly when conditioned on mutated receptors. We also explore the latent space learned by our generative model using sampling and interpolation techniques. This work opens the door for end-to-end prediction of stable bioactive molecules from protein structures with deep learning.

We generate 3D molecules conditioned on receptor binding sites by training a deep generative model on protein–ligand complexes. Our model uses the conditional receptor information to make chemically relevant changes to the generated molecules.     

Graph-based molecular alignment (GMA)

We describe a combined 2D/3D approach for the superposition of flexible chemical structures, which is based on recent progress in the efficient identification of common subgraphs and a gradient-based torsion space optimization algorithm. The simplicity of the approach is reflected in its generality and computational efficiency: the suggested approach neither requires precalculated statistics on the conformations of the molecules nor does it make simplifying assumptions on the topology of the molecules being compared. Furthermore, graph-based molecular alignment produces alignments that are consistent with the chemistry of the molecules as well as their general structure, as it depends on both the local connectivities between atoms and the overall topology of the molecules. We validate this approach on benchmark sets taken from the literature and show that it leads to good results compared to computationally and algorithmically more involved methods. The results suggest that, for most practical purposes, graph-based molecular alignment is a viable alternative to molecular field alignment with respect to structural superposition and leads to structures of comparable quality in a fraction of the time.     

Ultrafast vibrational spectroscopy of water and aqueous N-methylacetamide: Comparison of different electronic structure/molecular dynamics approaches

Kwac and Cho [J. Chem. Phys. 119, 2247 (2003)] have recently developed a combined electronic structure/molecular dynamics approach to vibrational spectroscopy in liquids. The method involves fitting ab initio vibrational frequencies for a solute in a cluster of solvent molecules to a linear combination of the electrostatic potentials on the solute atoms due to the charges on the solvent molecules. These authors applied their method to the N-methylacetamide-D/D(2)O system. We (S. A. Corcelli, C. P. Lawrence, and J. L. Skinner, [J. Chem. Phys. 120, 8107 (2004)]) have recently explored a closely related method, where instead of the electrostatic potential, the solute vibrational frequencies are fit to the components of the electric fields on the solute atoms due to the solvent molecules. We applied our method to the HOD/D(2)O and HOD/H(2)O systems. In order to make a direct comparison of these two approaches, in this paper we apply their method to the water system, and our method to the N-methylacetamide system. For the water system we find that the electric field method is superior to the potential approach, as judged by comparison with experiments for the absorption line shape. For the N-methylacetamide system the two methods are comparable.     

On the dynamics of the H+ +D2(v=0,j=0)-->HD+D + reaction: a comparison between theory and experiment

The H+ +D2(v=0,j=0)-->HD+D + reaction has been theoretically investigated by means of a time independent exact quantum mechanical approach, a quantum wave packet calculation within an adiabatic centrifugal sudden approximation, a statistical quantum model, and a quasiclassical trajectory calculation. Besides reaction probabilities as a function of collision energy at different values of the total angular momentum, J, special emphasis has been made at two specific collision energies, 0.1 and 0.524 eV. The occurrence of distinctive dynamical behavior at these two energies is analyzed in some detail. An extensive comparison with previous experimental measurements on the Rydberg H atom with D2 molecules has been carried out at the higher collision energy. In particular, the present theoretical results have been employed to perform simulations of the experimental kinetic energy spectra.     

A comprehensive evaluation of the kinetic method applied in the determination of the proton affinity of the nucleic acid molecules

The determination of proton affinity (PA) of 2'-deoxyadenosine (dA) is used as a case study for the evaluation of possible drawbacks in the determination of the PA of the nucleic acid molecules by the kinetic method. The observed Delta PA among the different values obtain for dA by applying this procedure in its different extensions was 0.64 Kcal/mol, which is within the uncertainties of any theoretical or experimental approach. It was demonstrated that nucleosides can be generally used as reference compounds to measure the PA of an unknown nucleoside. The evaluation of Delta Delta S value for two competing reaction channels taken by proton-bound heterodimers formed by two nucleosides provides clear information on the reference base which has to be discarded from the set of reference compounds used for the estimation of an unknown PA. The PA of dA obtained with the most elaborated kinetic method (237.00 +/- 0.07 kcal/mol) is consistent with the value of 237.0 kcal/mol obtained by a simple treatment of the relative intensities of the product ions generated by two couples of the proton bound dimers formed by the nucleoside and two reference amines. The kinetic method can be, therefore, confidently used to assess the proton affinity of the multifunctional molecules such as nucleosides and nucleobases.     

Interactions between block copolymers and single-walled carbon nanotubes in aqueous solutions: a small-angle neutron scattering study

The amphiphilic copolymers of the Pluronic family are known to be excellent dispersants for single-walled carbon nanotubes (SWCNT) in water, especially F108 and F127, which have rather long end-blocks of poly(ethylene oxide) (PEO). In this study, the structure of the CNT/polymer hybrid formed in water is evaluated by measurements of small-angle neutron scattering (SANS) with contrast variation, as supported by cryo-transmission electron microscopy (cryo-TEM) imaging. The homogeneous, stable, inklike dispersions exhibited very small isolated bundles of carbon nanotubes in cryo-TEM images. SANS experiments were conducted at different D(2)O/H(2)O content of the dispersing solvent. The data for both systems showed surprisingly minimal intensity values at 70% D(2)O solvent composition, which is much higher than the expected value of 17% D(2)O that is based on the scattering length density (SLD) of PEO. At this near match point, the data exhibited a q(-1) power law relation of intensity to the scattering vector (q), indicating rodlike entities. Two models are evaluated, as extensions to Pederson's block copolymer micelles models. One is loosely adsorbed polymer chains on a rodlike CNT bundle. In the other, the hydrophobic block is considered to form a continuous hydrated shell on the CNT surface, whereas the hydrophilic blocks emanate into the solvent. Both models were found to fit the experimental data reasonably well. The model fit required special considerations of the tight association of water molecules around PEO chains and slight isotopic selectivity.     

Ligand-based molecular modeling study on a chemically diverse series of cholecystokinin-B/gastrin receptor antagonists: generation of predictive model

Pharmacophore hypotheses were developed for six structurally diverse series of cholecystokinin-B/gastrin receptor (CCK-BR) antagonists. A training set consisting of 33 compounds was carefully selected. The activity spread of the training set molecules was from 0.1 to 2100 nM. The most predictive pharmacophore model (hypothesis 1), consisting of four features, namely, two hydrogen bond donors, one hydrophobic aliphatic, and one hydrophobic aromatic feature, had a correlation (r) of 0.884 and a root-mean-square deviation of 1.1526, and the cost difference between null cost and fixed cost was 81.5 bits. The model was validated on a test set consisting of six different series of 27 structurally diverse compounds and performed well in classifying active and inactive molecules correctly. This validation approach provides confidence in the utility of the predictive pharmacophore model developed in this work as a 3D query tool in the virtual screening of drug-like molecules to retrieve new chemical entities as potent CCK-BR antagonists. The model can also be used to predict the biological activities of compounds prior to their costly and time-consuming synthesis.     

Water solvent effects using continuum and discrete models: The nitromethane molecule,CH3NO2

The first three valence transitions of the two nitromethane conformers (CH₃NO₂) are two dark n → π* transitions and a very intense π → π* transition. In this work, these transitions in gas‐phase and solvated in water of both conformers were investigated theoretically. The polarizable continuum model (PCM), two conductor‐like screening (COSMO) models, and the discrete sequential quantum mechanics/molecular mechanics (S‐QM/MM) method were used to describe the solvation effect on the electronic spectra. Time dependent density functional theory (TDDFT), configuration interaction including all single substitutions and perturbed double excitations (CIS(D)), the symmetry‐adapted‐cluster CI (SAC‐CI), the multistate complete active space second order perturbation theory (CASPT2), and the algebraic‐diagrammatic construction (ADC(2)) electronic structure methods were used. Gas‐phase CASPT2, SAC‐CI, and ADC(2) results are in very good agreement with published experimental and theoretical spectra. Among the continuum models, PCM combined either with CASPT2, SAC‐CI, or B3LYP provided good agreement with available experimental data. COSMO combined with ADC(2) described the overall trends of the transition energy shifts. The effect of increasing the number of explicit water molecules in the S‐QM/MM approach was discussed and the formation of hydrogen bonds was clearly established. By including explicitly 24 water molecules corresponding to the complete first solvation shell in the S‐QM/MM approach, the ADC(2) method gives more accurate results as compared to the TDDFT approach and with similar computational demands. The ADC(2) with S‐QM/MM model is, therefore, the best compromise for accurate solvent calculations in a polar environment. © 2015 Wiley Periodicals, Inc.     

Calculation of free energies and chemical potentials for gas hydrates using Monte Carlo simulations

We describe a method for calculating free energies and chemical potentials for molecular models of gas hydrate systems using Monte Carlo simulations. The method has two components: (i) thermodynamic integration to obtain the water and guest molecule chemical potentials as functions of the hydrate occupancy; (ii) calculation of the free energy of the zero-occupancy hydrate system using thermodynamic integration from an Einstein crystal reference state. The approach is applicable to any classical molecular model of a hydrate. We illustrate the methodology with an application to the structure-I methane hydrate using two molecular models. Results from the method are also used to assess approximations in the van der Waals-Platteeuw theory and some of its extensions. It is shown that the success of the van der Waals-Platteeuw theory is in part due to a cancellation of the error arising from the assumption of a fixed configuration of water molecules in the hydrate framework with that arising from the neglect of methane-methane interactions.     

Dynamic Alignment of D2 Enhanced by Two Few-cycle Pulses

Dynamic alignment of D2 induced by two few-cycle pulses was investigated by solving the time-dependent Schr?odinger equation numerically based on a rigid rotor model. The results show that alignment of D2 can be enhanced by two few-cycle pulses compared with the level achievable by a single few-cycle pulse as long as the time delay between two pulses is chosen properly, and the pulse duration of two lasers plays an important role in the aligning process of D2 molecules.     

GPCR structure-based virtual screening approach for CB2 antagonist search

The potential for therapeutic specificity in regulating diseases has made cannabinoid (CB) receptors one of the most important G-protein-coupled receptor (GPCR) targets in search for new drugs. Considering the lack of related 3D experimental structures, we have established a structure-based virtual screening protocol to search for CB2 bioactive antagonists based on the 3D CB2 homology structure model. However, the existing homology-predicted 3D models often deviate from the native structure and therefore may incorrectly bias the in silico design. To overcome this problem, we have developed a 3D testing database query algorithm to examine the constructed 3D CB2 receptor structure model as well as the predicted binding pocket. In the present study, an antagonist-bound CB2 receptor complex model was initially generated using flexible docking simulation and then further optimized by molecular dynamic and mechanical (MD/MM) calculations. The refined 3D structural model of the CB2-ligand complex was then inspected by exploring the interactions between the receptor and ligands in order to predict the potential CB2 binding pocket for its antagonist. The ligand-receptor complex model and the predicted antagonist binding pockets were further processed and validated by FlexX-Pharm docking against a testing compound database that contains known antagonists. Furthermore, a consensus scoring (CScore) function algorithm was established to rank the binding interaction modes of a ligand on the CB2 receptor. Our results indicated that the known antagonists seeded in the testing database can be distinguished from a significant amount of randomly chosen molecules. Our studies demonstrated that the established GPCR structure-based virtual screening approach provided a new strategy with a high potential for in silico identifying novel CB2 antagonist leads based on the homology-generated 3D CB2 structure model.     

Prediction of hydrogen hydrate equilibrium by integrating ab initio calculations with statistical thermodynamics

This paper addresses a new calculation approach for the prediction of hydrogen hydrate equilibrium by introducing the concept of a single hydrogen cluster in one cavity. By integrating ab initio calculations with classical statistical thermodynamics, this approach enables the van der Waals model to predict the dissociation pressure of hydrogen hydrates. Compared to hydrates formed by light hydrocarbon gases, structure II (sII) hydrogen hydrates stably encage two and four hydrogen molecules in the small and large cavities, respectively. By treating two hydrogen molecules or four hydrogen molecules as one rigid body cluster, we determine ab initio binding energies between water molecules and hydrogen clusters at the MP2 level with the 6-31++G(2d,2p) basis set. These binding energies will be used to determine the parameters of the Exp-6 potential function from which the smooth cell potential and the Langmuir constant of each cluster are calculated. Then, the dissociation pressure is determined using the Zele-Lee-Holder cell distortion model: 105, 625, and 2000 bar at 150, 200, and 250 K, respectively.     

When is a molecule properly solvated by a continuum model or in a cluster ansatz? A first-principles simulation of alanine hydration

In order to test the validity of the cluster ansatz approach as well as of the continuum model approach and to learn about the solvation shell, we carried out first-principles molecular dynamics simulations of the alanine hydration. Our calculations contained one alanine molecule dissolved in 60 water molecules. Dipole moments of individual molecules were derived by means of maximally localized Wannier functions. We observed an average dipole moment of about 16.0 D for alanine and of about 3.3 D for water. In particular, the average water dipole moment in proximity of alanine's COO(-) group decayed continously with increasing distance, while, surprisingly, close to the CH3 and NH3+ group, the dipole moment first rose before its value dropped. In a cluster ansatz approach, we considered snapshots of alanine surrounded by different water molecule shells. The dipole moments from the cluster approaches utilizing both maximally localized Wannier functions as well as natural population analysis served to approximate the dipole moments of the total trajectory. Sufficient convergence of the cluster ansatz approach is found for either of the two solvent shells around the polar groups and one solvent shell around the apolar groups or two solvent shells around the polar groups surrounded by a dieletric continuum.     

An efficient fragment-based approach for predicting the ground-state energies and structures of large molecules

An efficient fragment-based approach for predicting the ground-state energies and structures of large molecules at the Hartree-Fock (HF) and post-HF levels is described. The physical foundation of this approach is attributed to the "quantum locality" of the electron correlation energy and the HF total energy, which is revealed by a new energy decomposition analysis of the HF total energy proposed in this work. This approach is based on the molecular fractionation with conjugated caps (MFCC) scheme (Zhang, D. W.; Zhang, J. Z. H. J. Chem. Phys. 2003, 119, 3599), by which a macromolecule is partitioned into various capped fragments and conjugated caps formed by two adjacent caps. We find that the MFCC scheme, if corrected by the interaction between non-neighboring fragments, can be used to predict the total energy of large molecules only from energy calculations on a series of small subsystems. The approach, named as energy-corrected MFCC (EC-MFCC), computationally achieves linear scaling with the molecular size. Our test calculations on a broad range of medium- and large molecules demonstrate that this approach is able to reproduce the conventional HF and second-order Moller-Plesset perturbation theory (MP2) energies within a few millihartree in most cases. With the EC-MFCC optimization algorithm described in this work, we have obtained the optimized structures of long oligomers of trans-polyacetylene and BN nanotubes with up to about 400 atoms, which are beyond the reach of traditional computational methods. In addition, the EC-MFCC approach is also applied to estimate the heats of formation for a series of organic compounds. This approach provides an appealing approach alternative to the traditional additivity rules based on either bond or group contributions for the estimation of thermochemical properties.