Blind prediction of toluene/water partition coefficients using COSMO-RS: results from the SAMPL9 challenge.

Accurately predicting partition coefficients log P is crucial for reducing costs and accelerating drug design as it provides valuable information about the bioavailability, pharmacokinetics, and toxicity of different drug candidates. However, the performance of the existing methods is ambiguous, making it unclear whether these methods can be effectively utilized in drug discovery. To assess the performance of these methods, a series of SAMPL challenges have been conducted over the past few years, aiming to enable the development and validation of predictive models. In this study, we present two independent contributions to the SAMPL9 challenge for predicting the toluene/water partition coefficients for 16 molecules. Both submissions, A and B, use the COSMO-RS approach, albeit in slightly different procedures, to compute the transfer free energies from water to toluene of the molecules presented in the challenge, and consequently, their corresponding log P values. Based on the results, COSMO-RS submission A achieves the top position with an R2 value of 0.93 while it ranks second in terms of root-mean-square error (RMSE) with a value of 1.23 log P units. COSMO-RS submission B achieves an R2 value of 0.83 and an RMSE value of 1.48 log P units. Following the challenge, we predict the log P values using a neural network model, which was pre-trained on COSMO-RS data achieving an R2 of 0.92 and RMSE of 1.04 log P units. Compared to previous SAMPL challenges, all contributions displayed large deviations in predicting the toluene/water partition coefficient. These large deviations emphasize that further research is needed to develop accurate and reliable methods for modeling solvent effects on small molecule transfer-free energies.

Prediction of acid pKa values in the solvent acetone based on COSMO-RS.

In this contribution we extent the use of the conductor-like screening model for realistic solvation (COSMO-RS) to the prediction of pKa values in acetone, a commonly used dipolar aprotic solvent. For this, we calculated the Gibbs free energy of dissociation of 120 organic acids (nine acrylic acids, 87 benzoic acids, nine phenols, and 15 benzenesulfonamides) using COSMO-RS at the two levels BP-TZVP and BP-TZVPD-FINE and determined the parameters for a linear free energy relation for the pKa prediction by performing linear fits to experimental values. Our results suggest that the data set dissects into two groups, with the phenols being different from the other three subsets. The acrylic and benzoic acids and the sulfonamides can be treated together and yield an excellent linear correlation ( r2>0.95 ) with an RMSD of only ~0.3. The slope is found to be significantly smaller than the theoretical value ( 1/RTln10 ), only 45% of it, which is in accordance with findings in the literature. The phenols, however, while similarly well correlated in their own subset with an RMSD of 1.7-1.9, exhibit a slope larger than one. We discuss both a higher uncertainty in the reference values as well as physical origins as possible reasons.

TURBOMOLE: Modular program suite for ab initio quantum-chemical and condensed-matter simulations.

TURBOMOLE is a collaborative, multi-national software development project aiming to provide highly efficient and stable computational tools for quantum chemical simulations of molecules, clusters, periodic systems, and solutions. The TURBOMOLE software suite is optimized for widely available, inexpensive, and resource-efficient hardware such as multi-core workstations and small computer clusters. TURBOMOLE specializes in electronic structure methods with outstanding accuracy-cost ratio, such as density functional theory including local hybrids and the random phase approximation (RPA), GW-Bethe-Salpeter methods, second-order Møller-Plesset theory, and explicitly correlated coupled-cluster methods. TURBOMOLE is based on Gaussian basis sets and has been pivotal for the development of many fast and low-scaling algorithms in the past three decades, such as integral-direct methods, fast multipole methods, the resolution-of-the-identity approximation, imaginary frequency integration, Laplace transform, and pair natural orbital methods. This review focuses on recent additions to TURBOMOLE's functionality, including excited-state methods, RPA and Green's function methods, relativistic approaches, high-order molecular properties, solvation effects, and periodic systems. A variety of illustrative applications along with accuracy and timing data are discussed. Moreover, available interfaces to users as well as other software are summarized. TURBOMOLE's current licensing, distribution, and support model are discussed, and an overview of TURBOMOLE's development workflow is provided. Challenges such as communication and outreach, software infrastructure, and funding are highlighted.

Development of new auxiliary basis functions of the Karlsruhe segmented contracted basis sets including diffuse basis functions (def2-SVPD, def2-TZVPPD, and def2-QVPPD) for RI-MP2 and RI-CC calculations.

We report optimized auxiliary basis sets for use with the Karlsruhe segmented contracted basis sets including moderately diffuse basis functions (Rappoport and Furche, J. Chem. Phys., 2010, 133, 134105) in resolution-of-the-identity (RI) post-self-consistent field (post-SCF) computations for the elements H-Rn (except lanthanides). The errors of the RI approximation using optimized auxiliary basis sets are analyzed on a comprehensive test set of molecules containing the most common oxidation states of each element and do not exceed those of the corresponding unaugmented basis sets. During these studies an unsatisfying performance of the def2-SVP and def2-QZVPP auxiliary basis sets for Barium was found and improved sets are provided. We establish the versatility of the def2-SVPD, def2-TZVPPD, and def2-QZVPPD basis sets for RI-MP2 and RI-CC (coupled-cluster) energy and property calculations. The influence of diffuse basis functions on correlation energy, basis set superposition error, atomic electron affinity, dipole moments, and computational timings is evaluated at different levels of theory using benchmark sets and showcase examples.

Heuristic control of kinetic energy in dynamic reaction coordinate calculations.

For the understanding and prediction of chemical reactions, detailed knowledge of the minimum energy path between reactants and transition state is of utmost importance. Stewart et al. (J. Comput. Chem. 1987, 8, 1117) proposed the usage of molecular trajectories calculated from Newton's equations of motion for an efficient reaction path following. Two operational modes are possible thereby: intrinsic (IRC) and dynamic reaction coordinate calculations (DRC). The technical difference between these modes is that in an IRC calculation the kinetic energy of the nuclei is quenched while the total energy is conserved in DRC calculations. In this work, a heuristic control methodology of atomic kinetic energies in DRC calculations using fuzzy logic is proposed. A diversified test set of 10 reactions has been collected to examine the performance of this approach. Fuzzy rule-based models are found to be a convenient way to make the determination of accessible paths of chemical reactions computationally efficient.

thermocalc - a poor man's approach to computational thermochemistry.

We present thermocalc, a Perl module to perform the automated calculation of atomization energies and heats of formation for lists of molecules. The methods used are based on density functional theory and second-order perturbation theory to ensure that data sets of medium sized to large molecules can be run at reasonable throughput rates. The quantum chemical calculations are performed using the program package TURBOMOLE in a three-step protocol. In a first step, a pre-optimization of the structure and a zero-point energy calculation are performed. As second step, a geometry optimization is being carried out, and the last step is a single point energy calculation. The level of theory used in the different steps can be modified by the user to allow for customized protocols. The performance of example protocols is investigated on different test sets of molecules. In the course of this work, a simple, but efficient one-parameter correction term based on the shared electron numbers has been developed, which reduces the error of calculated heats of formation significantly.

Polarization charge densities provide a predictive quantification of hydrogen bond energies.

A systematic density functional theory based study of hydrogen bond energies of 2465 single hydrogen bonds has been performed. In order to be closer to liquid phase conditions, different from the usual reference state of individual donor and acceptor molecules in vacuum, the reference state of donors and acceptors embedded in a perfect conductor as simulated by the COSMO solvation model has been used for the calculation of the hydrogen bond energies. The relationship between vacuum and conductor reference hydrogen bond energies is investigated and interpreted in the light of different physical contributions, such as electrostatic energy and dispersion. A very good correlation of the DFT/COSMO hydrogen bond energies with conductor polarization charge densities of separated donor and acceptor atoms was found. This provides a method to predict hydrogen bond strength in solution with a root mean square error of 0.36 kcal mol(-1) relative to the quantum chemical dimer calculations. The observed correlation is broadly applicable and allows for a predictive quantification of hydrogen bonding, which can be of great value in many areas of computational, medicinal and physical chemistry.

The accuracy of dipole moments from spin-component scaled CC2 in ground and electronically excited states.

The accuracy of dipole moments calculated from wave function methods based on second-order perturbation theory is investigated in the ground and electronically excited states. Results from the approximate coupled-cluster singles-and-doubles model, CC2, Møller-Plesset perturbation theory, MP2, and the algebraic diagrammatic construction through second-order, ADC(2), are discussed together with the spin-component scaled and the scaled opposite-spin variants of these methods. The computed dipole moments show a very good correlation with data from high-resolution spectroscopy. Compared to the unscaled methods, the spin-component scaling increases the accuracy of the results and improves the robustness of the calculations. An accuracy about 0.2 to 0.1 D in the ground state and about 0.3 to 0.2 D in the electronically excited states can be achieved with these approaches.

TmoleX--a graphical user interface for TURBOMOLE.

We herein present the graphical user interface (GUI) TmoleX for the quantum chemical program package TURBOMOLE. TmoleX allows users to execute the complete workflow of a quantum chemical investigation from the initial building of a structure to the visualization of the results in a user friendly graphical front end. The purpose of TmoleX is to make TURBOMOLE easy to use and to provide a high degree of flexibility. Hence, it should be a valuable tool for most users from beginners to experts. The program is developed in Java and runs on Linux, Windows, and Mac platforms. It can be used to run calculations on local desktops as well as on remote computers.

Recyclization rate of a photocleaved peptide from multiscale simulation.

Unprecedented insight into the phototriggered unfolding of a polypeptide has been gained from a multiscale simulation connecting nonadiabatic ab initio molecular dynamics to classical molecular dynamics in a three-stage manner. An intramolecular H-transfer mechanism that saturates one of the S. radicals of the cleaved S-S bridge and thus prevents recyclization has been observed. This chemical quenching mechanism may be the key to resolving the controversy surrounding the S-S reformation rates.

Preferential pathways for light-trapping involving beta-ligated chlorophylls.

The magnesium atom of chlorophylls (Chls) is always five- or six-coordinated within chlorophyll-protein complexes which are the main light-harvesting systems of plants, algae and most photosynthetic bacteria. Due to the presence of stereocenters and the axial ligation of magnesium the two faces of Chls are diastereotopic. It has been previously recognized that the alpha-configuration having the magnesium ligand on the opposite face of the 17-propionic acid moiety is more frequently encountered and is more stable than the more seldom beta-configuration that has the magnesium ligand on the same face [T.S. Balaban, P. Fromme, A.R. Holzwarth, N. Kraubeta, V.I. Prokhorenko, Relevance of the diastereotopic ligation of magnesium atoms in chlorophylls in Photosystem I, Biochim. Biophys. Acta (Bioenergetics), 1556 (2002) 197-207; T. Oba, H. Tamiaki, Which side of the pi-macrocycle plane of (bacterio)chlorophylls is favored for binding of the fifth ligand? Photosynth. Res. 74 (2002) 1-10]. In photosystem I only 14 Chls out of a total of 96 are in a beta-configuration and these occupy preferential positions around the reaction center. We have now analyzed the alpha/beta dichotomy in the homodimeric photosystem II based on the 2.9 A resolution crystal structure [A. Guskov, J. Kern, A. Gabdulkhakov, M. Broser, A. Zouni, W. Saenger, Cyanobacterial photosystem II at 2.9 A resolution: role of quinones, lipids, channels and chloride, Nature Struct. Mol. Biol. 16 (2009) 334-342] and find that out of 35 Chls in each monomer only 9 are definitively in the beta-configuration, while 4 are uncertain. Ab initio calculations using the approximate coupled-cluster singles-and-doubles model CC2 [O. Christiansen, H. Koch, P. Jørgensen, The second-order approximate coupled cluster singles and doubles model CC2, Chem. Phys. Lett. 243 (1995) 409-418] now correctly predict the absorption spectra of Chls a and b and conclusively show for histidine, which is the most frequent axial ligand of magnesium in chlorophyll-protein complexes, that only slight differences (<4 nm) are encountered between the alpha- and beta-configurations. Significant red shifts (up to 50 nm) can, however, be encountered in excitonically coupled beta-beta-Chl dimers. Surprisingly, in both photosystems I and II very similar "special" beta-beta dimers are encountered at practically the same distances from P700 and P680, respectively. In purple bacteria LH2, the B850 ring is composed exclusively of such tightly coupled beta-bacteriochlorophylls a. A statistical analysis of the close contacts with the protein matrix (<5 A) shows significant differences between the alpha- and beta-configurations and the subunit providing the axial magnesium ligand. The present study allows us to conclude that the excitation energy transfer in light-harvesting systems, from a peripheral antenna towards the reaction center, may follow preferential pathways due to structural reasons involving beta-ligated Chls.

Benchmarking the performance of spin-component scaled CC2 in ground and electronically excited states.

A generalization of the spin-component scaling and scaled opposite-spin modifications of second-order Møller-Plesset perturbation theory to the approximate coupled-cluster singles-and-doubles model CC2 (termed SCS-CC2 and SOS-CC2) is discussed and a preliminary implementation of ground and excited state energies and analytic gradients is reported. The computational results for bond distances, harmonic frequencies, adiabatic and 0-0 excitation energies are compared with experimental results to benchmark their performance. It is found that both variants of the spin-scaling increase the robustness of CC2 against strong correlation effects and lead for this method even to somewhat larger improvements than those observed for second-order Møller-Plesset perturbation theory. The spin-component scaling also enhances systematically the accuracy of CC2 for 0-0 excitation energies for pi --> pi* and n --> pi* transitions, if geometries are determined at the same level.

On the internal rotations in p-cresol in its ground and first electronically excited states.

The overall rotation and internal rotation of p-cresol (4-methyl-phenol) has been studied by comparison of the microwave spectrum with accurate ab initio calculations using the principal axis method in the electronic ground state. Both internal rotations, the torsions of the methyl and the hydroxyl groups relative to the aromatic ring, have been investigated. The internal rotation of the hydroxyl group can be approximately described as the motion of a symmetrical rotor on an asymmetric frame. For the methyl group it has been found that the potential barrier hindering its internal rotation is very small with the first two nonvanishing Fourier coefficients of the potential V(3) and V(6) in the same order of magnitude. Different splittings of b-type transitions for the A and E species of the methyl torsion indicate a top-top interaction between both internal rotors through the benzene ring. An effective coupling potential for the top-top interaction could be estimated. The hindering barriers of the hydroxyl and methyl rotation have been calculated using second-order Moller-Plesset perturbation theory and the approximate coupled-cluster singles-and-doubles model (CC2) in the ground state and using CC2 and the algebraic diagrammatic construction through second order in the first electronically excited state. The results are in excellent agreement with the experimental values.

Intramolecular charge-transfer mechanism in quinolidines: the role of the amino twist angle.

Quantum-chemical calculations with the approximate coupled-cluster singles-and-doubles model CC2 have been carried out for 1-tert-butyl-6-cyano-1,2,3,4-tetrahydroquinoline (NTC6). For this molecule dual fluorescence was experimentally observed, raising the discussion about the importance of the amino twist angle for this process. The calculations suggest that both the ground state and the normal fluorescent state are significantly twisted by 30 degrees -40 degrees and that the molecule is flexible enough to move into an even stronger twisted conformation (60 degrees -70 degrees ) in its intramolecular charge-transfer (ICT) state which is responsible for the anomalous fluorescence band. Such a conformation both minimizes the total energy in the S1 state and maximizes the dipole moment. The barrier from the normal fluorescent state to the ICT state region is very small. Comparison to the situation in the 1-methyl-derivative NMC6 suggests that a large alkyl substituent makes the preferably planar normal fluorescent state energetically unfavorable compared to the ICT state and thus promotes the occurrence of dual fluorescence.

Microwave and theoretical investigation of the internal rotation in m-cresol.

The microwave spectrum of m-cresol (3-methylphenol) has been investigated using a molecular beam Fourier transform microwave spectrometer in the frequency range from 3 to 26.5 GHz. The rotation of the hydroxy group into two different unequal energetic minima leads to different spectra for the syn- and anticonformers. Because of a high potential barrier both conformers can be analyzed independently. The methyl group is undergoing an almost free internal rotation which is only hindered by small barriers and splits the vibrational ground state in two states of internal rotation denoted as A and E species. The spacing between the species is found to be up to 10 GHz. The potential for the internal rotation can be determined from the spectra and analyzed in terms of the Fourier components V3 and V6. For syn-m-cresol these parameters were determined as V3=673(3) GHz and V6=-335(24) GHz and for anti-m-cresol V3=95(5) GHz and V6=-416(46) GHz. The barriers to internal rotation were furthermore calculated with second-order Moller-Plesset perturbation theory and second-order coupled-cluster singles- and-doubles model (CC2) in the electronic ground state and with CC2 in the first excited state. The CC2 method is found to be an appropriate method to calculate potential barriers in electronic excited states of such compounds.

Distributed memory parallel implementation of energies and gradients for second-order Møller-Plesset perturbation theory with the resolution-of-the-identity approximation.

We present a parallel implementation of second-order Møller-Plesset perturbation theory with the resolution-of-the-identity approximation (RI-MP2). The implementation is based on a recent improved sequential implementation of RI-MP2 within the Turbomole program package and employs the message passing interface (MPI) standard for communication between distributed memory nodes. The parallel implementation extends the applicability of canonical MP2 to considerably larger systems. Examples are presented for full geometry optimizations with up to 60 atoms and 3300 basis functions and MP2 energy calculations with more than 200 atoms and 7000 basis functions.