Importance of dispersion and electron correlation in ab initio protein folding

Dispersion Forces-Driven Hierarchical Assembly of Protein-Like Lanthanide Octamers and Emergent CPL

Hierarchical self-assembly driven by non-covalent interactions is a prevalent strategy employed by nature to construct sophisticated biomacromolecules, such as proteins. However, the construction of protein-like superstructures that rely on weaker dispersion forces-driven hierarchical assembly remains largely unexplored. Here, we report the first example of dispersion forces driving the high-order assembly of the lanthanide trinuclear circular helicate [HNEt₃]₃[Eu₃(LL)₆] (ΔΔΔ-1) into a protein-like lanthanide octamer ((ΔΔΔ-1)₈-2). Within the octamer, the forty-eight (48) menthol groups on the ligands and eighty-four (84) 1,4-dioxane solvent molecules contribute to enhanced dispersion forces through conformational adaptation and size-matching effects. These enhanced dispersion forces not only drive the formation of the hierarchical superstructure but also result in a four-level chirality transfer from the menthol to the octamer. Benefiting from the homochirality of Eu3+, the octamer is endowed the strong circularly polarized emission (|glum|=0.34, Φoverall=41 %). This understanding of how dispersion forces drive hierarchical self-assembly provides a foundation for the directed fabrication of more fascinating superstructures.

Accurate Neural Network Fine-Tuning Approach for Transferable Ab Initio Energy Prediction across Varying Molecular and Crystalline Scales

Existing machine learning models attempt to predict the energies of large molecules by training small molecules, but eventually fail to retain high accuracy as the errors increase with system size. Through an orbital pairwise decomposition of the correlation energy, a pretrained neural network model on hundred-scale data containing small molecules is demonstrated to be sufficiently transferable for accurately predicting large systems, including molecules and crystals. Our model introduces a residual connection to explicitly learn the pairwise energy corrections, and employs various low-rank retraining techniques to modestly adjust the learned network parameters. We demonstrate that with as few as only one larger molecule retraining the base model originally trained on only small molecules of (H2O)6, the MP2 correlation energy of the large liquid water (H2O)64 in a periodic supercell can be predicted at chemical accuracy. Similar performance is observed for large protonated clusters and periodic poly glycine chains. A demonstrative application is presented to predict the energy ordering of symmetrically inequivalent sublattices for distinct hydrogen orientations in the ice XV phase. Our work represents an important step forward in the quest for cost-effective, highly accurate and transferable neural network models in quantum chemistry, bridging the electronic structure patterns between small and large systems.

Substituent and Heteroatom Effects on π-π Interactions: Evidence That Parallel-Displaced π-Stacking is Not Driven by Quadrupolar Electrostatics

Stacking interactions are a recurring motif in supramolecular chemistry and biochemistry, where a persistent theme is a preference for parallel-displaced aromatic rings rather than face-to-face π-stacking. This is typically explained in terms of quadrupole-quadrupole interactions between the arene moieties but that interpretation is inconsistent with accurate calculations, which reveal that the quadrupolar picture is qualitatively wrong. At typical π-stacking distances, quadrupolar electrostatics may differ in sign from an exact calculation based on charge densities of the interacting arenes. We apply symmetry-adapted perturbation theory to dimers composed of substituted benzene and various aromatic heterocycles, which display a wide range of electrostatic interactions, and we investigate the interplay of Pauli repulsion, dispersion, and electrostatics as it pertains to parallel-displaced π-stacking. Profiles of energy components along cofacial slip-stacking coordinates support a prominent role for the "van der Waals model" (dispersion in competition with Pauli repulsion), even for polar monomers where electrostatic interactions are significant. While electrostatic interactions are necessary to explain the optimal face-to-face π-stacking distance and to account for the relative orientation of one polar arene with respect to another, we find no evidence to support continued invocation of quadrupolar electrostatics as a basis for π-stacking. Our results suggest that a driving force for offset-stacking exists even in the absence of electrostatic interactions. Consequently, tuning electrostatics via functionalization does not guarantee that slip-stacking can be avoided. This has implications for rational design of soft materials and other supramolecular architectures.

Modeling Intermolecular Interactions with Exchange-Hole Dipole Moment Dispersion Corrections to Neural Network Potentials

Neural network potentials (NNPs) are an innovative approach for calculating the potential energy and forces of a chemical system. In principle, these methods are capable of modeling large systems with an accuracy approaching that of a high-level ab initio calculation, but with a much smaller computational cost. Due to their training to density-functional theory (DFT) data and neglect of long-range interactions, some classes of NNPs require an additional term to include London dispersion physics. In this Perspective, we discuss the requirements for a dispersion model for use with an NNP, focusing on the MLXDM (Machine Learned eXchange-Hole Dipole Moment) model developed by our groups. This model is based on the DFT-based XDM dispersion correction, which calculates interatomic dispersion coefficients in terms of atomic moments and polarizabilities, both of which can be approximated effectively using neural networks.

Partition analysis of dipole moments in solution applied to functional groups in polypeptide motifs

A partition analysis based on segments is developed for density functional theory defining solute dipole moments of functional groups, and the corresponding induced solvent dipoles representing solvent screening. The accuracy of dipoles from the fragment molecular orbital method is evaluated in comparison to unfragmented values. The analysis is applied to evaluate dipole moments of side chains, amino and carbonyl groups in common polypeptide motifs, α-helixes, β-turns, and random coils in solution. The membrane domain of the ATP synthase (1B9U) is analyzed, revealing the effect of the bend splitting of the α-helix into two pieces.

Use of caps in the auxiliary basis set formulation of the fragment molecular orbital method

An auxiliary polarization formulation of the fragment molecular orbital (FMO) method is developed, combining a basis set correction computed for capped isolated fragments with a polarization obtained from uncapped fragments. For a set of organic and inorganic test systems, it is shown that the total energy and atomic charges are accurately reproduced with respect to full unfragmented calculations. It is demonstrated that the method is accurate for computing electronic excited states. The developed approach is applied to rank the isomers of chignolin from experimental NMR data (PDB: 1UAO) according to their relative energy. Contributions of polarization and basis set effects to pair interactions between fragments are elucidated.

100 Years of the Lennard-Jones Potential

It is now 100 years since Lennard-Jones published his first paper introducing the now famous potential that bears his name. It is therefore timely to reflect on the many achievements, as well as the limitations, of this potential in the theory of atomic and molecular interactions, where applications range from descriptions of intermolecular forces to molecules, clusters, and condensed matter.

Comprehensive Basis-Set Testing of Extended Symmetry-Adapted Perturbation Theory and Assessment of Mixed-Basis Combinations to Reduce Cost

Hybrid or "extended" symmetry-adapted perturbation theory (XSAPT) replaces traditional SAPT's treatment of dispersion with better performing alternatives while at the same time extending two-body (dimer) SAPT to a many-body treatment of polarization using a self-consistent charge embedding procedure. The present work presents a systematic study of how XSAPT interaction energies and energy components converge with respect to the choice of Gaussian basis set. Errors can be reduced in a systematic way using correlation-consistent basis sets, with aug-cc-pVTZ results converged within <0.1 kcal/mol. Similar (if slightly less systematic) behavior is obtained using Karlsruhe basis sets at much lower cost, and we introduce new versions with limited augmentation that are even more efficient. Pople-style basis sets, which are more efficient still, often afford good results if a large number of polarization functions are included. The dispersion models used in XSAPT afford much faster basis-set convergence as compared to the perturbative description of dispersion in conventional SAPT, meaning that "compromise" basis sets (such as jun-cc-pVDZ) are no longer required and benchmark-quality results can be obtained using triple-ζ basis sets. The use of diffuse functions proves to be essential, especially for the description of hydrogen bonds. The "δ(Hartree-Fock)" correction for high-order induction can be performed in double-ζ basis sets without significant loss of accuracy, leading to a mixed-basis approach that offers 4× speedup over the existing (cubic scaling) XSAPT approach.

Predicting and Understanding Non-Covalent Interactions Using Novel Forms of Symmetry-Adapted Perturbation Theory

Although sometimes derided as "weak" interactions, non-covalent forces play a critical role in ligand binding and crystal packing and in determining the conformational landscape of flexible molecules. Symmetry-adapted perturbation theory (SAPT) provides a framework for accurate ab initio calculation of intermolecular interactions and furnishes a natural decomposition of the interaction energy into physically meaningful components: semiclassical electrostatics (rigorously obtained from monomer charge densities), Pauli or steric repulsion, induction (including both polarization and charge transfer), and dispersion. This decomposition helps to foster deeper understanding of non-covalent interactions and can be used to construct transferable, physics-based force fields. Separability of the SAPT interaction energy also provides the flexibility to construct composite methods, a feature that we exploit to improve the description of dispersion interactions. These are challenging to describe accurately because they arise from nonlocal electron correlation effects that appear for the first time at second order in perturbation theory but are not quantitatively described at that level.As with all quantum-chemical methods, a major limitation of SAPT is nonlinear scaling of the computational cost with respect to system size. This cost can be significantly mitigated using "SAPT0(KS)", which incorporates monomer electron correlation by means of Kohn-Sham (KS) molecular orbitals from density functional theory (DFT), as well as by an "extended" theory called XSAPT, developed by the authors. XSAPT generalizes traditional dimer SAPT to many-body systems, so that a ligand-protein interaction (for example) can be separated into contributions from individual amino acids, reducing the cost of the calculation below that of even supramolecular DFT while retaining the accuracy of high-level ab initio quantum chemistry.This Account provides an overview of the SAPT0(KS) approach and the XSAPT family of methods. Several low-cost variants are described that provide accuracy approaching that of the best ab initio benchmarks yet are affordable enough to tackle ligand-protein binding and sizable host-guest complexes. These variants include SAPT+aiD, which uses ab initio atom-atom dispersion potentials ("+aiD") in place of second-order SAPT dispersion, and also SAPT+MBD, which incorporates many-body dispersion (MBD) effects that are important in the description of nanoscale materials. Applications to drug binding highlight the size-extensive nature of dispersion, which is not a weak interaction in large systems. Other applications highlight how a physics-based analysis can sometimes upend conventional wisdom regarding intermolecular forces. In particular, careful reconsideration of π-π interactions makes clear that the quadrupolar electrostatics (or "Hunter-Sanders") model of π-π stacking should be replaced by a "van der Waals model" in which conformational preferences arise from a competition between dispersion and Pauli repulsion. Our analysis also suggests that molecular shape, rather than aromaticity per se, is the key factor driving strong stacking interactions. Looking forward, we anticipate that XSAPT-based methods can play a role in screening of drug candidates and in materials design.

Molecular conversion of MIG6 hotspot-3 peptide from the nonbinder to a moderate binder of HER2 by rational design of an orthogonal interaction system at the HER2-peptide interface

Human epidermal growth factor receptor 2 (HER2) has been established as an approved druggable target for the treatment of patients with diverse gynecological tumors such as ovarian, cervical and breast cancers. The mitogen-inducible gene 6 (MIG6) protein is a negative regulator of HER2 signaling by using its Seg1 segment to disrupt the allosteric dimerization of HER2 kinase domain. Previous studies found that the Seg1 adopts three separated hotspots to interact with the HER2 dimerization interface, in which the third hotspot (H3) is located at the core region of the interface but its derived H3 peptide (356PKYVS360) and Tyr358Phe mutant (356PKFVS360) cannot bind effectively to the interface in an independent manner. In this study, we demonstrate that the H3 peptide can be converted from nonbinder to a moderate binder of HER2 by just adding an orthogonal noncovalent interaction system (X⋯O┄H) between a halogen bond (X⋯O) and a hydrogen bond (H┄O) involving peptide Phe358 residue and HER2 Val948/Trp951 residues. High-level calculations are utilized to rigorously characterize and rationally design the X⋯O┄H system, which is then optimized with different halogen atoms and at different substituting positions. It is revealed that there is a synergistic effect between the X⋯O and H┄O of the orthogonal interaction system; formation of the halogen bond can enhance the interaction strength of the hydrogen bond. In silico analysis and in vitro assay reach a consistence that Br-substitution at the m-position of peptide Phe358 phenyl moiety is the best choice that can render strong interaction for the X⋯O┄H system, which also makes the peptide 'bindable' to HER2 kinase domain, while F/Cl/I-substitution at the same position can only improve the peptide affinity moderately or modestly. In contrast, the Br-substitution at the o- and p-positions of peptide Phe358 phenyl moiety cannot define effective X⋯O┄H interaction and thus does not confer additional affinity to the HER2-peptide complex.

Intrinsic relative preference profile of pan-kinase inhibitor drug staurosporine towards the clinically occurring gatekeeper mutations in Protein Tyrosine Kinases

Protein tyrosine kinases (PTKs) have been recognized as the attractive druggable targets of various diseases including cancer. However, many PTKs are clinically observed to establish a gatekeeper mutation in the peripheral hinge section of active site, which plays a primary role in development of acquired drug resistance to kinase inhibitors. The natural product Staurosporine, an ATP-competitive reversible pan-kinase inhibitor, has been found to exhibit wild type-sparing selectivity for some PTK gatekeeper mutants. In this study, totally 23 acquired drug-resistant gatekeeper mutations harbored on 17 PTKs involved in diverse cancers were curated, from which only five amino acid types, namely Thr, Met, Val, Leu and Ile, were observed at both wild-type and mutant residues of these clinically occurring gatekeeper sites. Here, an integrative strategy that combined molecular modeling and kinase assay was described to systematically investigate the relative preference of Staurosporine towards the five gatekeeper amino acid types in real kinase context and in a psendokinase model. A kinase-free, intrinsic relative preference profile of Staurosporine to gatekeeper amino acids was created: (dispreferred) Thr⊳Val⊳Ile⊳Leu⊳Met (preferred). It is found that kinase context has no essential effect on the profile; different kinases and even psendokinase can obtain a consistent conclusion for the preference order. Theoretically, we can use the profile to predict Staurosporine response to any gatekeeper mutation between the five amino acid types in any PTK. Structural and energetic analyses revealed that the multiple-aromatic ring system of Staurosporine can form multiple noncovalent interactions with the weakly polar side chain of Met and can pack tightly or moderately against the nonpolar side chains of Val, Ile and Leu, thus stabilizing the kinase-inhibitor system (ΔU < 0), whereas the polar side chain of Thr may cause unfavorable electronegative and solvent effects with the aromatic electrons of Staurosporine, thus destabilizing the system (ΔU > 0).

Combining the Fragmentation Approach and Neural Network Potential Energy Surfaces of Fragments for Accurate Calculation of Protein Energy

Accurate and efficient all-atom quantum mechanical (QM) calculations for biomolecules still present a challenge to computational physicists and chemists. In this study, an extensible generalized molecular fractionation with a conjugate caps method combined with neural networks (NN-GMFCC) is developed for efficient QM calculation of protein energy. In the NN-GMFCC scheme, the total energy of a given protein is calculated by taking a proper combination of the high-precision neural network potential energies of all capped residues and overlapping conjugate caps. In addition, the two-body interaction energies of residue pairs are calculated by molecular mechanics (MM). With reference to the GMFCC/MM calculation at the ωB97XD/6-31G* level, the overall mean unsigned errors of the energy deviations and atomic force root-mean-squared errors calculated by NN-GMFCC are only 2.01 kcal/mol and 0.68 kcal/mol/Å, respectively, for 14 proteins (containing up to 13,728 atoms). Meanwhile, the NN-GMFCC approach is about 4 orders of magnitude faster than the GMFCC/MM method. The NN-GMFCC method could be systematically improved by inclusion of two-body QM interaction and multibody electronic polarization effect. Moreover, the NN-GMFCC approach can also be applied to other macromolecular systems such as DNA/RNA, and it is capable of providing a powerful and efficient approach for exploration of structures and functions of proteins with QM accuracy.

Analyzing Interactions with the Fragment Molecular Orbital Method

Basic concepts in the analysis of binding using the fragment molecular orbital method are discussed at length: polarization, desolvation, and interaction. The components in the pair interaction energy decomposition analysis are introduced, and the analysis is illustrated for a water dimer and a protein-ligand complex.

Tetrel Interactions from an Interacting Quantum Atoms Perspective

Tetrel bonds, the purportedly non-covalent interaction between a molecule that contains an atom of group 14 and an anion or (more generally) an atom or molecule with lone electron pairs, are under intense scrutiny. In this work, we perform an interacting quantum atoms (IQA) analysis of several simple complexes formed between an electrophilic fragment (A) (CH₃F, CH₄, CO₂, CS₂, SiO₂, SiH₃F, SiH₄, GeH₃F, GeO₂, and GeH₄) and an electron-pair-rich system (B) (NCH, NCO-, OCN-, F-, Br-, CN-, CO, CS, Kr, NC-, NH₃, OC, OH₂, SH-, and N₃-) at the aug-cc-pvtz coupled cluster singles and doubles (CCSD) level of calculation. The binding energy ( E bind AB ) is separated into intrafragment and inter-fragment components, and the latter in turn split into classical and covalent contributions. It is shown that the three terms are important in determining E bind AB , with absolute values that increase in passing from electrophilic fragments containing C, Ge, and Si. The degree of covalency between A and B is measured through the real space bond order known as the delocalization index ( δ AB ). Finally, a good linear correlation is found between δ AB and E xc AB , the exchange correlation (xc) or covalent contribution to E bind AB .

Structure-based screening and validation of potential dengue virus inhibitors through classical and QM/MM affinity estimation

The recurrent outbreaks of dengue virus around the globe represent a huge challenge for governments and public health organizations. With the rapid growth and ease of transportation, dengue disease continues to spread, placing more of the world population under constant threat. Despite decades of research efforts, no effective small molecule antivirals are available against dengue virus. With the efficacy of the recently developed vaccine to be determined, there is an urgent unmet need for small molecule dengue virus treatments. In the current study, we employed state-of-the-art molecular modelling simulations to identify novel inhibitors of the dengue virus envelope protein. The binding modes of all compounds within the conserved β-OctylGlucoside (β-OG) pocket were studied using a combination of docking, molecular dynamics simulations and binding free energy calculations. Here, we describe ten new compounds that significantly reduce production of dengue virus as determined using standard cell-based virological assays. Moreover, we present a comprehensive structural analysis of the identified hits, focusing on their electrostatic and lipophilic binding energy contributions. Finally, we highlight the effect of the desolvation penalty in limiting the activity of some of these compounds. The data presented here paves the way toward rationally designing selective and potent novel inhibitors against dengue virus.

Fragment Molecular Orbital Calculations with Implicit Solvent Based on the Poisson-Boltzmann Equation: II. Protein and Its Ligand-Binding System Studies

In this study, the electronic properties of bioactive proteins were analyzed using an ab initio fragment molecular orbital (FMO) methodology in solution: coupling with an implicit solvent model based on the Poisson-Boltzmann surface area called as FMO-PBSA. We investigated the solvent effects on practical and heterogeneous targets with uneven exposure to solvents unlike deoxyribonucleic acid analyzed in our recent study. Interfragment interaction energy (IFIE) and its decomposition analyses by FMO-PBSA revealed solvent-screening mechanisms that affect local stability inside ubiquitin protein: the screening suppresses excessiveness in bare charge-charge interactions and enables an intuitive IFIE analysis. The electrostatic character and associated solvation free energy also give consistent results as a whole to previous studies on the explicit solvent model. Moreover, by using the estrogen receptor alpha (ERα) protein bound to ligands, we elucidated the importance of specific interactions that depend on the electric charge and activatability as agonism/antagonism of the ligand while estimating the influences of the implicit solvent on the ligand and helix-12 bindings. The predicted ligand-binding affinities of bioactive compounds to ERα also show a good correlation with their in vitro activities. The FMO-PBSA approach would thus be a promising tool both for biological and pharmaceutical research targeting proteins.

Effect of Electron Correlation on Intermolecular Interactions: A Pair Natural Orbitals Coupled Cluster Based Local Energy Decomposition Study

The development of post-Hartree-Fock (post-HF) energy decomposition schemes that are able to decompose the HF and correlation components of the interaction energy into chemically meaningful contributions is a very active field of research. One of the challenges is to provide a clear-cut quantification to the elusive London dispersion component of the intermolecular interaction. London dispersion is well-known to be a pure correlation effect, and as such it is not properly described by mean field theories. In this context, we have recently developed the local energy decomposition (LED) analysis, which provides a chemically meaningful decomposition of the interaction energy between two or more fragments computed at the domain-based local pair natural orbitals coupled cluster (DLPNO-CCSD(T)) level of theory. In this work, this scheme is used in conjunction with other interpretation tools to study a series of molecular adducts held together by intermolecular interactions of different natures. The HF and correlation components of the interaction energy are thus decomposed into a series of chemically meaningful contributions. Emphasis is placed on discussing the physical effects associated with the inclusion of electron correlation. It is found that four distinct physical effects can contribute to the magnitude of the correlation part of intermolecular binding energies (Δ E_int^C): (i) London dispersion, (ii) the correlation correction to the reference induction energy, (iii) the correlation correction to the electron sharing process, and (iv) the correlation correction to the permanent electrostatics. As expected, the largest contribution to the correlation binding energy of neutral, apolar molecules is London dispersion, as in the argon dimer case. In contrast, the correction for the HF induction energy dominates Δ E_int^C in systems in which an apolar molecule interacts with charged or strongly polar species, as in Ar-Li⁺. This effect has its origin in the systematic underestimation of polarizabilities at the HF level of theory. For similar reasons, electron sharing largely contributes to the correlation binding energy of covalently bound molecules, as in the beryllium dimer case. Finally, the correction for HF permanent electrostatics significantly contributes to Δ E_int^C in molecules with strong dipoles, such as water and hydrogen fluoride dimers. This effect originates from the characteristic overestimation of dipole moments at the HF level of theory, leading in some cases to positive Δ E_int^C values. Our results are apparently in contrast to the widely accepted view that Δ E_int^C is typically dominated by London dispersion, at least, in the strongly interacting region. Clearly, post-HF energy decomposition schemes are very powerful tools to analyze, categorize, and understand the various contributions to the intermolecular interaction energy. Hopefully, this will eventually lead to insights that are helpful in designing systems with tailored properties. All analysis tools presented in this work will be available free of charge in the next release of the ORCA program package.

The fragment molecular orbital method combined with density-functional tight-binding and the polarizable continuum model

The electronic gap in proteins is analyzed in detail, and it is shown that FMO-DFTB/PCM is efficient and accurate in describing the molecular structure of proteins in solution.

A simple and efficient dispersion correction to the Hartree-Fock theory (2): Incorporation of a geometrical correction for the basis set superposition error

One of the most challenging problems in computer-aided drug discovery is the accurate prediction of the binding energy between a ligand and a protein. For accurate estimation of net binding energy ΔEbind in the framework of the Hartree-Fock (HF) theory, it is necessary to estimate two additional energy terms; the dispersion interaction energy (Edisp) and the basis set superposition error (BSSE). We previously reported a simple and efficient dispersion correction, Edisp, to the Hartree-Fock theory (HF-Dtq). In the present study, an approximation procedure for estimating BSSE proposed by Kruse and Grimme, a geometrical counterpoise correction (gCP), was incorporated into HF-Dtq (HF-Dtq-gCP). The relative weights of the Edisp (Dtq) and BSSE (gCP) terms were determined to reproduce ΔEbind calculated with CCSD(T)/CBS or /aug-cc-pVTZ (HF-Dtq-gCP (scaled)). The performance of HF-Dtq-gCP (scaled) was compared with that of B3LYP-D3(BJ)-bCP (dispersion corrected B3LYP with the Boys and Bernadi counterpoise correction (bCP)), by taking ΔEbind (CCSD(T)-bCP) of small non-covalent complexes as 'a golden standard'. As a critical test, HF-Dtq-gCP (scaled)/6-31G(d) and B3LYP-D3(BJ)-bCP/6-31G(d) were applied to the complex model for HIV-1 protease and its potent inhibitor, KNI-10033. The present results demonstrate that HF-Dtq-gCP (scaled) is a useful and powerful remedy for accurately and promptly predicting ΔEbind between a ligand and a protein, albeit it is a simple correction procedure.

Non-pairwise additivity of the leading-order dispersion energy

The leading-order (i.e., dipole-dipole) dispersion energy is calculated for one-dimensional (1D) and two-dimensional (2D) infinite lattices, and an infinite 1D array of infinitely long lines, of doubly occupied locally harmonic wells. The dispersion energy is decomposed into pairwise and non-pairwise additive components. By varying the force constant and separation of the wells, the non-pairwise additive contribution to the dispersion energy is shown to depend on the overlap of density between neighboring wells. As well separation is increased, the non-pairwise additivity of the dispersion energy decays. The different rates of decay for 1D and 2D lattices of wells is explained in terms of a Jacobian effect that influences the number of nearest neighbors. For an array of infinitely long lines of wells spaced 5 bohrs apart, and an inter-well spacing of 3 bohrs within a line, the non-pairwise additive component of the leading-order dispersion energy is -0.11 kJ mol(-1) well(-1), which is 7% of the total. The polarizability of the wells and the density overlap between them are small in comparison to that of the atomic densities that arise from the molecular density partitioning used in post-density-functional theory (DFT) damped dispersion corrections, or DFT-D methods. Therefore, the nonadditivity of the leading-order dispersion observed here is a conservative estimate of that in molecular clusters.

Quality Assessment of Predicted Protein Models Using Energies Calculated by the Fragment Molecular Orbital Method

Protein structure prediction directly from sequences is a very challenging problem in computational biology. One of the most successful approaches employs stochastic conformational sampling to search an empirically derived energy function landscape for the global energy minimum state. Due to the errors in the empirically derived energy function, the lowest energy conformation may not be the best model. We have evaluated the use of energy calculated by the fragment molecular orbital method (FMO energy) to assess the quality of predicted models and its ability to identify the best model among an ensemble of predicted models. The fragment molecular orbital method implemented in GAMESS was used to calculate the FMO energy of predicted models. When tested on eight protein targets, we found that the model ranking based on FMO energies is better than that based on empirically derived energies when there is sufficient diversity among these models. This model diversity can be estimated prior to the FMO energy calculations. Our result demonstrates that the FMO energy calculated by the fragment molecular orbital method is a practical and promising measure for the assessment of protein model quality and the selection of the best protein model among many generated.

Fragment quantum mechanical calculation of proteins and its applications

Conspectus The desire to study molecular systems that are much larger than what the current state-of-the-art ab initio or density functional theory methods could handle has naturally led to the development of novel approximate methods, including semiempirical approaches, reduced-scaling methods, and fragmentation methods. The major computational limitation of ab initio methods is the scaling problem, because the cost of ab initio calculation scales nth power or worse with system size. In the past decade, the fragmentation approach based on chemical locality has opened a new door for developing linear-scaling quantum mechanical (QM) methods for large systems and for applications to large molecular systems such as biomolecules. The fragmentation approach is highly attractive from a computational standpoint. First, the ab initio calculation of individual fragments can be conducted almost independently, which makes it suitable for massively parallel computations. Second, the electron properties, such as density and energy, are typically combined in a linear fashion to reproduce those for the entire molecular system, which makes the overall computation scale linearly with the size of the system. In this Account, two fragmentation methods and their applications to macromolecules are described. They are the electrostatically embedded generalized molecular fractionation with conjugate caps (EE-GMFCC) method and the automated fragmentation quantum mechanics/molecular mechanics (AF-QM/MM) approach. The EE-GMFCC method is developed from the MFCC approach, which was initially used to obtain accurate protein-ligand QM interaction energies. The main idea of the MFCC approach is that a pair of conjugate caps (concaps) is inserted at the location where the subsystem is divided by cutting the chemical bond. In addition, the pair of concaps is fused to form molecular species such that the overcounted effect from added concaps can be properly removed. By introducing the electrostatic embedding field in each fragment calculation and two-body interaction energy correction on top of the MFCC approach, the EE-GMFCC method is capable of accurately reproducing the QM molecular properties (such as the dipole moment, electron density, and electrostatic potential), the total energy, and the electrostatic solvation energy from full system calculations for proteins. On the other hand, the AF-QM/MM method was used for the efficient QM calculation of protein nuclear magnetic resonance (NMR) parameters, including the chemical shift, chemical shift anisotropy tensor, and spin-spin coupling constant. In the AF-QM/MM approach, each amino acid and all the residues in its vicinity are automatically assigned as the QM region through a distance cutoff for each residue-centric QM/MM calculation. Local chemical properties of the central residue can be obtained from individual QM/MM calculations. The AF-QM/MM approach precisely reproduces the NMR chemical shifts of proteins in the gas phase from full system QM calculations. Furthermore, via the incorporation of implicit and explicit solvent models, the protein NMR chemical shifts calculated by the AF-QM/MM method are in excellent agreement with experimental values. The applications of the AF-QM/MM method may also be extended to more general biological systems such as DNA/RNA and protein-ligand complexes.

Density functional theory based study on cis-trans isomerism of the amide bond in homodimers of β(2,3)- and β(3)-substituted homoproline

Preference for a cis/trans peptide bond between residues of dipeptides formed by substituted β(2,3) (I) and β(3) (II) homoproline is investigated using density functional theory (DFT). Potential energy surfaces for monomer and linear dimers are explored at the B3LYP/6-31G(d,p) level of theory. Minimum energy conformations of the dipeptides are optimized using B3LYP, PBE1PBE, B97D, and M06-2X functionals at the 6-31G(d,p) level of basis set in both the gas phase and solvent phase. The relative free energy difference between the selected conformations is marginal. Results obtained using the functionals M06-2X and B97D on dimers of I and II, respectively, agree with experimental results. The lowest energy conformations predicted by B97D/6-31G(d,p) and M06-2X/6-31G(d,p) levels of theory show greater relative MP2 correlation energy. Dipeptides of I with hydrophilic substituents show preference for a trans peptide bond. Support for cis/trans isomerism in dimers of I with hydrophobic substituents comes from potential energy surfaces and free energy data. Although dipeptides of II with hydrophilic substituents show preference for cis peptide bond, the dipeptides with hydrophobic substituent prefer trans bond.

QM and QM/MM simulations of proteins

Molecular dynamics simulations of biomolecules have matured into powerful tools of structural biology. In addition to the commonly used empirical force field potentials, quantum mechanical descriptions are gaining popularity for structure optimization and dynamic simulations of peptides and proteins. In this chapter, we introduce methodological developments such as the QM/MM framework and linear-scaling QM that make efficient calculations on large biomolecules possible. We identify the most common scenarios in which quantum descriptions of peptides and proteins are employed, such as structural refinement, force field development, treatment of unusual residues, and predicting spectroscopic and exited state properties. The benefits and shortcomings of QM potentials, in comparison to classical force fields, are discussed, with special emphasis on the sampling problems of protein conformational space. Finally, recent examples of QM/MM calculations in light-sensitive membrane proteins illustrate typical applications of the reviewed methods.

Synthesis and evaluation of new guanidine-thiourea organocatalyst for the nitro-Michael reaction: Theoretical studies on mechanism and enantioselectivity

A new guanidine-thiourea organocatalyst has been developed and applied as bifunctional organocatalyst in the Michael addition reaction of diethyl malonate to trans-β-nitrostyrene. Extensive DFT calculations, including solvent effects and dispersion corrections, as well as ab initio calculations provide a plausible description of the reaction mechanism.

Combined QM/MM (ONIOM) and QSAR approach to the study of complex formation of matrix metalloproteinase‑9 with a series of biphenylsulfonamides−LERE-QSAR analysis (V)

We previously proposed a novel QSAR (quantitative structure-activity relationship) procedure called LERE (linear expression by representative energy terms)-QSAR involving molecular calculations such as an ab initio fragment molecular orbital ones. In the present work, we applied LERE-QSAR to complex formation of matrix metalloproteinase-9 (MMP-9) with a series of substituted biphenylsulfonamides. The results shows that the overall free-energy change accompanying complex formation is due to predominantly the contribution from the electrostatic interaction with the zinc atom in the active site of MMP-9. Carbonic anhydrase (CA) belongs to the zinc-containing protease family. In contrast to the current case of MMP-9, the overall free-energy change during complex formation of CA with a series of benzenesulfonamides is due to the contributions from the solvation and dissociation free-energy changes, as previously reported. Comparison of the two sets of results indicates quantitative differences in the relative contributions of free-energy components to the overall free-energy change between the two data sets, corresponding with those in the respective classical QSAR equations. The LERE-QSAR procedure was demonstrated to quantitatively reveal differences in the binding mechanisms between the two cases involving similar but different zinc-containing proteins at the electronic and atomic levels.

Computational and experimental studies of the interaction between phospho-peptides and the C-terminal domain of BRCA1

The C-terminal domain of BRCA1(BRCT) is involved in the DNA repair pathway by recognizing the pSXXF motif in interacting proteins. It has been reported that short peptides containing this motif bind to BRCA1(BRCT) in the micromolar range with high specificity. In this work, the binding of pSXXF peptides has been studied computationally and experimentally in order to characterize their interaction with BRCA1(BRCT). Elucidation of the contacts that drive the protein-ligand interaction is critical for the development of high affinity small-molecule BRCA1 inhibitors. Molecular dynamics (MD) simulations revealed the key role of threonine at the peptide P+2 position in providing structural rigidity to the ligand in the bound state. The mutation at P+1 had minor effects. Peptide extension at the N-terminal position with the naphthyl amino acid exhibited a modest increase in binding affinity, what could be explained by the dispersion interaction of the naphthyl side-chain with a hydrophobic patch. Three in silico end-point methods were considered for the calculation of binding free energy. The Molecular Mechanics Poisson-Boltzmann Surface Area and the Solvated Interaction Energy methods gave reasonable agreement with experimental data, exhibiting a Pearlman predictive index of 0.71 and 0.78, respectively. The MM-quantum mechanics-surface area method yielded improved results, which was characterized by a Pearlman index of 0.78. The correlation coefficients were 0.59, 0.61 and 0.69, respectively. The ability to apply a QM level of theory within an end-point binding free energy protocol may provide a way for a consistent improvement of accuracy in computer-aided drug design.

Model for the fast estimation of basis set superposition error in biomolecular systems

Basis set superposition error (BSSE) is a significant contributor to errors in quantum-based energy functions, especially for large chemical systems with many molecular contacts such as folded proteins and protein-ligand complexes. While the counterpoise method has become a standard procedure for correcting intermolecular BSSE, most current approaches to correcting intramolecular BSSE are simply fragment-based analogues of the counterpoise method which require many (two times the number of fragments) additional quantum calculations in their application. We propose that magnitudes of both forms of BSSE can be quickly estimated by dividing a system into interacting fragments, estimating each fragment's contribution to the overall BSSE with a simple statistical model, and then propagating these errors throughout the entire system. Such a method requires no additional quantum calculations, but rather only an analysis of the system's interacting fragments. The method is described herein and is applied to a protein-ligand system, a small helical protein, and a set of native and decoy protein folds.