COSMOperm: Mechanistic Prediction of Passive Membrane Permeability for Neutral Compounds and Ions and Its pH Dependence.

We present a new and entirely mechanistic COSMOperm method to predict passive membrane permeabilities for neutral compounds, as well as anions and cations. The COSMOperm approach is based on compound-specific free energy profiles within a membrane of interest from COSMO-RS (conductor-like screening model for realistic solvation) calculations. These are combined with membrane layer-specific diffusion coefficients, for example, in the water phase, the polar head groups, and the alkyl tails of biochemical phospholipid bilayers. COSMO-RS utilizes first-principle quantum chemical structures and physically sound intermolecular interactions (electrostatic, hydrogen bond, and van der Waals). For this reason, it is unbiased toward different application scenarios, such as in cosmetics and industrial chemical or pharmaceutical industries. A fully predictive calculation of passive permeation through phospholipid bilayer membranes results in a performance of r2 = 0.92; rmsd = 0.90 log10 units for neutral compounds and anions, as compared to gold standard black lipid membrane experiments. It will be demonstrated that new membrane types can be generated by the related COSMOplex method and directly used for permeability studies by COSMOperm.

COSMO-RS based predictions for the SAMPL6 logP challenge.

Within the framework of the 6th physical property blind challenge (SAMPL6) the authors have participated in predicting the octanol-water partition coefficients (logP) for several small drug like molecules. Those logP values where experimentally known by the organizers but only revealed after the submissions of the predictions. Two different sets of predictions were submitted by the authors, both based on the COSMOtherm implementation of COSMO-RS theory. COSMOtherm predictions using the FINE parametrization level (hmz0n) obtained the highest accuracy among all submissions as measured by the root mean squared error. COSMOquick predictions using a fast algorithm to estimate σ-profiles and an a posterio machine learning correction on top of the COSMOtherm results (3vqbi) scored 3rd out of 91 submissions. Both results underline the high quality of COSMO-RS derived molecular free energies in solution.

Prediction of Partition Coefficients of Environmental Toxins Using Computational Chemistry Methods.

The partitioning of compounds between aqueous and other phases is important for predicting toxicity. Although thousands of octanol-water partition coefficients have been measured, these represent only a small fraction of the anthropogenic compounds present in the environment. The octanol phase is often taken to be a mimic of the inner parts of phospholipid membranes. However, the core of such membranes is typically more hydrophobic than octanol, and other partition coefficients with other compounds may give complementary information. Although a number of (cheap) empirical methods exist to compute octanol-water (log k OW) and hexadecane-water (log k HW) partition coefficients, it would be interesting to know whether physics-based models can predict these crucial values more accurately. Here, we have computed log k OW and log k HW for 133 compounds from seven different pollutant categories as well as a control group using the solvation model based on electronic density (SMD) protocol based on Hartree-Fock (HF) or density functional theory (DFT) and the COSMO-RS method. For comparison, XlogP3 (log k OW) values were retrieved from the PubChem database, and KowWin log k OW values were determined as well. For 24 of these compounds, log k OW was computed using potential of mean force (PMF) calculations based on classical molecular dynamics simulations. A comparison of the accuracy of the methods shows that COSMO-RS, KowWin, and XlogP3 all have a root-mean-square deviation (rmsd) from the experimental data of ≈0.4 log units, whereas the SMD protocol has an rmsd of 1.0 log units using HF and 0.9 using DFT. PMF calculations yield the poorest accuracy (rmsd = 1.1 log units). Thirty-six out of 133 calculations are for compounds without known log k OW, and for these, we provide what we consider a robust prediction, in the sense that there are few outliers, by averaging over the methods. The results supplied may be instrumental when developing new methods in computational ecotoxicity. The log k HW values are found to be strongly correlated to log k OW for most compounds.

Ab initio prediction of structuring/mesoscale inhomogeneities in surfactant-free microemulsions and hydrogen-bonding-free microemulsions.

In this paper, we consider the influence of H-bond donor and acceptor functionalities on the formation of mesoscale inhomogeneities in ternary systems. It was found that hydrogen-bonding re-enforces such structures, but is not necessarily a prerequisite for the occurrence of mesoscale, microemulsion-like structuring in ternary surfactant-free microemulsions (SFME) and consequently, hydrogen-bonding-free microemulsions (HBFME) exist. The evaluated ternary systems were investigated by means of dynamic light scattering (DLS) and computer-based calculation methods. Theoretical COSMO-RS based calculations were applied to provide an explanation for different hydrotropic efficiencies, and COSMOplex calculations were used to predict and evaluate the propensity of the molecules to form mesoscale structures in SFME and HBFME. Microemulsion-like fluctuations could be observed in the COSMOplex simulations and correlate fairly well with the appearance of mesoscopic structures observed in SFME and HBFME, although the free energy differences in the formation of aggregate structures in the investigated systems are very small, in the range of 0.05 kcal mol-1.

COSMOplex: self-consistent simulation of self-organizing inhomogeneous systems based on COSMO-RS.

During the past 20 years, the efficient combination of quantum chemical calculations with statistical thermodynamics by the COSMO-RS method has become an important alternative to force-field based simulations for the accurate prediction of free energies of molecules in liquid systems. While it was originally restricted to homogeneous liquids, it later has been extended to the prediction of the free energy of molecules in inhomogeneous systems such as micelles, biomembranes, or liquid interfaces, but these calculations were based on external input about the structure of the inhomogeneous system. Here we report the rigorous extension of COSMO-RS to a self-consistent prediction of the structure and the free energies of molecules in self-organizing inhomogeneous systems. This extends the application range to many new areas, such as the prediction of micellar structures and critical micelle concentrations, finite loading effects in micelles and biomembranes, the free energies and structure of liquid interfaces, microemulsions, and many more related topics, which often are of great practical importance.

A refined cavity construction algorithm for the conductor-like screening model.

A cavity construction algorithm based on the triangulation of an iso-surface is introduced as a new standard for dielectric continuum solvation calculations with the Conductor-like Screening Model COSMO. It overcomes deficiencies which have become apparent for the original COSMO standard cavity, especially in concave regions of the molecular shaped cavity. The new standard, called FINE Cavity, is described in this article with several application examples. The earlier COSMO cavity construction algorithms are described for comparison. © 2018 Wiley Periodicals, Inc.

Prediction of cyclohexane-water distribution coefficients with COSMO-RS on the SAMPL5 data set.

The Conductor-Like-Screening-Model for Real Solvents (COSMO-RS) method has been used for the blind prediction of cyclohexane-water distribution coefficients logD within the SAMPL challenge. The partition coefficient logP of the neutral species was calculated first and then corrected for dissociation or protonation, as appropriate for acidic or basic solutes, to obtain the cyclohexane-water logD. Using the latest version of the COSMOtherm implementation, this approach in combination with a rigorous conformational sampling yielded a predictive accuracy of 2.11 log units (RMSD) for the 53 compounds of the blind prediction dataset. By that it was the most accurate of all contest submissions and it also achieved the best rank order. The RMSD mainly arises from a group of outliers in the negative logD range, which at least partly may arise from dimerization or other experimental problems coming up for very polar molecules in very non-polar solvents.

Computational Screening of Drug Solvates.

PURPOSE: Solvates are mainly undesired by-products during the pharmaceutical development of new drugs. In addition, solvate formation may also distort solubility measurements. The presented study introduces a simple computational approach that allows for the identification of drug solvent pairs which most likely form crystalline solid phases. METHODS: The mixing enthalpy as a measure for drug-solvent complementarity is obtained by computational liquid phase thermodynamics (COSMO-RS theory). In addition a few other simple descriptors were taking into account describing the shape and topology of the drug and the solvent. Using an extensive dataset of drug solvent pairs a simple and statistically robust model is developed which allows for a rough assessment of a solvent's ability to form a solvate. RESULTS: Similar to the related issue of cocrystal screening, the mixing (or excess) enthalpy of the subcooled liquid mixture of the drug-solvent pair proves to be an important quantity controlling solvate formation. Due to the fact that many solvates form inclusion compounds, the solvent shape is another important factor influencing solvate formation. Solvates forming channel-like voids in the solid state are predicted less well. CONCLUSION: The approach ranks any drug-solvent pair that forms a solvate before any non-solvate by a probability of about 81% (AUC = 0.81), giving a significant advantage over any trial and error approach. Hence it can help to identify suitable solvent candidates early in the drug development process.

Quantum chemical insights into the dependence of porphyrin basicity on the meso-aryl substituents: thermodynamics, buckling, reaction sites and molecular flexibility.

The chemical and sensing properties of porphyrins are frequently tuned via the introduction of peripheral substituents. In the context of the exceptionally fast second protonation step in the case of 5,10,15,20-tetraphenylporphyrin (TPP), as compared to porphin and 5,10,15,20-tetramesitylporphyrin (TMesP), we investigated the macrocycle-substituent interactions of these three porphyrin derivatives in detail. Using quantum chemical thermodynamics calculations, the analysis of geometric structures, torsional profiles, electrostatic potential distributions, and particularly the analysis of molecular flexibilities via ab initio molecular dynamics simulations, we obtained a comprehensive picture of the reactivities of the studied porphyrins and how these are influenced by the meso-substituents. As compared to porphin and TMesP the second protonation of TPP is energetically more favorable and is particularly energetically comparable to its first protonation, instead of being significantly less favorable like in the case of porphyrin and TMesP. Additionally, the second TPP protonation is facilitated by an interplay between out-of-plane (oop) distortion of the protonation site and a pronounced electrostatic binding spot at the protonation site. Furthermore, the second protonation is particularly facilitated in the case of TPP by the large oop-flexibility of the diprotonated species as unraveled by ab initio molecular dynamics simulations.

Solubility prediction, solvate and cocrystal screening as tools for rational crystal engineering.

OBJECTIVES: The fact that novel drug candidates are becoming increasingly insoluble is a major problem of current drug development. Computational tools may address this issue by screening for suitable solvents or by identifying potential novel cocrystal formers that increase bioavailability. In contrast to other more specialized methods, the fluid phase thermodynamics approach COSMO-RS (conductor-like screening model for real solvents) allows for a comprehensive treatment of drug solubility, solvate and cocrystal formation and many other thermodynamics properties in liquids. This article gives an overview of recent COSMO-RS developments that are of interest for drug development and contains several new application examples for solubility prediction and solvate/cocrystal screening. METHODS: For all property predictions COSMO-RS has been used. The basic concept of COSMO-RS consists of using the screening charge density as computed from first principles calculations in combination with fast statistical thermodynamics to compute the chemical potential of a compound in solution. KEY FINDING: The fast and accurate assessment of drug solubility and the identification of suitable solvents, solvate or cocrystal formers is nowadays possible and may be used to complement modern drug development. Efficiency is increased by avoiding costly quantum-chemical computations using a database of previously computed molecular fragments. SUMMARY: COSMO-RS theory can be applied to a range of physico-chemical properties, which are of interest in rational crystal engineering. Most notably, in combination with experimental reference data, accurate quantitative solubility predictions in any solvent or solvent mixture are possible. Additionally, COSMO-RS can be extended to the prediction of cocrystal formation, which results in considerable predictive accuracy concerning coformer screening. In a recent variant costly quantum chemical calculations are avoided resulting in a significant speed-up and ease-of-use.

Calculation of solvation free energies with DCOSMO-RS.

The concept of dielectric continuum models has turned out to be very fruitful for the qualitative description of solvation effects in quantum chemical calculations, although from a theoretical perspective its basis is questionable, at least if applied to polar solvents, because the electrostatic nearest neighbor interactions in polar solvents are much too strong to be described by macroscopic dielectric continuum theory. On the basis of this insight, the Conductorlike Screening Model for Realistic Solvation (COSMO-RS) had been developed, which gives a thermodynamically consistent, quantitative description of solvation effects in polar and nonpolar solvents, even in mixtures and at variable temperature, starting from quantum chemical calculations of solute and solvent molecules embedded in a virtual conductor (COSMO). Though COSMO-RS usually only requires quantum chemical calculations in the conductor and thus does not allow for studying of the concrete solvent influence on the solute electron density, the direct COSMO-RS (DCOSMO-RS) has been introduced, which uses the σ-potential, i.e., a solvent specific response function provided by COSMO-RS, as a replacement of the conductor or dielectric response employed in continuum solvation models. In this article we describe the current status of DCOSMO-RS and demonstrate the performance of the DCOSMO-RS approach for the prediction of free energies of solvation.

Prediction of Phospholipid-Water Partition Coefficients of Ionic Organic Chemicals Using the Mechanistic Model COSMOmic.

The partition coefficient of chemicals from water to phospholipid membrane, K(lipw), is of central importance for various fields. For neutral organic molecules, log K(lipw) correlates with the log of bulk solvent-water partition coefficients such as the octanol-water partition coefficient. However, this is not the case for charged compounds, for which a mechanistic modeling approach is highly necessary. In this work, we extend the model COSMOmic, which adapts the COSMO-RS theory for anisotropic phases and has been shown to reliably predict K(lipw) for neutral compounds, to the use of ionic compounds. To make the COSMOmic model applicable for ionic solutes, we implemented the internal membrane dipole potential in COSMOmic. We empirically optimized the potential with experimental K(lipw) data of 161 neutral and 75 ionic compounds, yielding potential shapes that agree well with experimentally determined potentials from the literature. This model refinement has no negative effect on the prediction accuracy of neutral compounds (root-mean-square error, RMSE = 0.62 log units), while it highly improves the prediction of ions (RMSE = 0.70 log units). The refined COSMOmic is, to our knowledge, the first mechanistic model that predicts K(lipw) of both ionic and neutral species with accuracies better than 1 log unit.

Prediction of free energies of hydration with COSMO-RS on the SAMPL4 data set.

The COSMO-RS method has been used for the prediction of free energies of hydration on a dataset of 47 complex multifunctional compounds considered in the SAMPL4 challenge. Straight application of the COSMOtherm software with the parameterization C21_0108 yields a predictive accuracy of 1.46 kcal/mol root mean square error overall and 1.18 kcal/mol if a single dominant outlier is removed.

Interpretation of experimental hydrogen-bond enthalpies and entropies from COSMO polarisation charge densities.

In this work, experimental hydrogen-bond (HB) enthalpies measured in previous works for a wide range of acceptor molecules in dilute mixtures of 4-fluorophenol in non-polar solvents are quantified from COSMO polarisation charge densities σ of HB acceptors (HBA). As well as previously demonstrated for quantum chemically calculated HB enthalpies, a good correlation of the experimental data with the polarisation charge densities is observed, covering an extended range of HBA (O, N, S, π systems and halogens) ranging from very weak to strong hydrogen bonds. Furthermore, for the first time, a quantitative analysis of experimental HB entropies is performed for such a chemical diversity of HBA. A good quantification of these entropies is achieved using the polarisation charge density σ as a descriptor in combination with the logarithm of a directional partition function Ω(HB). This partition function covers the directional and multiplicity entropy of HBA and is based on the σ-proportional HB enthalpy expression taken from COSMO-RS. As a result, the experimental HB enthalpies and free energies of the ~300 HB complexes are quantified with an accuracy of ~2 kJ mol(-1) based on COSMO polarisation charge densities.

Rational coformer or solvent selection for pharmaceutical cocrystallization or desolvation.

It is demonstrated that the fluid-phase thermodynamics theory conductor-like screening model for real solvents (COSMO-RS) as implemented in the COSMOtherm software can be used for accurate and efficient screening of coformers for active pharmaceutical ingredient (API) cocrystallization. The excess enthalpy, H(ex) , between an API-coformer mixture relative to the pure components reflects the tendency of those two compounds to cocrystallize. Thus, predictive calculations may be performed with decent effort on a large set of molecular data in order to identify potentially new cocrystal systems. In addition, it is demonstrated that COSMO-RS theory allows reasonable ranking of coformers for API solubility improvement. As a result, experiments may be focused on those coformers, which have an increased probability of cocrystallization, leading to the largest improvement of the API solubility. In a similar way as potential coformers are identified for cocrystallization, solvents that do not tend to form solvates may be determined based on the highest H(ex) s with the API. The approach was successfully tested on tyrosine kinase inhibitor axitinib, which has a propensity to form relatively stable solvated structures with the majority of common solvents, as well as on thiophanate-methyl and thiophanate-ethyl benzimidazole fungicides, which form channel solvates.

COSMOsar3D: molecular field analysis based on local COSMO σ-profiles.

The COSMO surface polarization charge density σ resulting from quantum chemical calculations combined with a virtual conductor embedding has been widely proven to be a very suitable descriptor for the quantification of interactions of molecules in liquids. In a preceding paper, grid-based local histograms of σ have been introduced in the COSMOsim3D method, resulting in a novel 3D-molecular similarity measure and going along with a novel property-based molecular alignment method. In this paper, we introduce under the name COSMOsar3D the usage of the resulting array of local σ-profiles as a novel set of molecular interaction fields for 3D-QSAR, containing all information required for quantifying the virtual ligand-receptor interactions, including desolvation. In contrast to currently used molecular interaction fields, we provide a theoretical rationale that the logarithmic binding constants of ligands should be a linear function of the array of local σ-profiles. This makes them especially suitable for linear regression analysis methods such as PLS. We demonstrate that the usage of local σ-profiles in molecular field analysis inverts the role of ligands and receptor; while conventional 3D-QSAR considers the virtual receptor in potential energy fields provided by the ligands, our COSMOsar3D approach corresponds to the calculation of the free energy of the ligands in a virtual free energy field provided by the receptor. First applications of the COSMOsar3D method are presented, which demonstrate its ability to yield robust and predictive models that seem to be superior to the models generated on the basis of conventionally used molecular fields.

COSMOsim3D: 3D-similarity and alignment based on COSMO polarization charge densities.

COSMO σ-surfaces resulting from quantum chemical calculations of molecules in a simulated conductor, and their histograms, the so-called σ-profiles, are widely proven to provide a very suitable and almost complete basis for the description of molecular interactions in condensed systems. The COSMOsim method therefore introduced a global measure of molecular similarity on the basis of similarity of σ-profiles, but it had the disadvantage of neglecting the 3D distribution of molecular polarities, which is crucially determining all ligand-receptor binding. This disadvantage is now overcome by COSMOsim3D, which is a logical and physically sound extension of the COSMOsim method, which uses local σ-profiles on a spatial grid. This new method is used to measure intermolecular similarities on the basis of the 3D representation of the surface polarization charge densities σ of the target and the probe molecule. The probe molecule is translated and rotated in space in order to maximize the sum of local σ-profile similarities between target and probe. This sum, the COSMOsim3D similarity, is a powerful descriptor of ligand similarity and allows for a good discrimination between bioisosters and random pairs. Validation experiments using about 600 pharmacological activity classes in the MDDR database are given. Furthermore, COSMOsim3D represents a unique and very robust method for a field-based ligand-ligand alignment.

Prediction of free energies of hydration with COSMO-RS on the SAMPL3 data set.

The COSMO-RS method has been used for the prediction of free energies of hydration on the dataset of 36 chlorinated ethanes, biphenyls and dioxins considered in the SAMPL3 challenge. Straight application of the latest version of the COSMOtherm software yields an overall predictive accuracy of 1.05 kcal/mol (RMSE). The predictions for the chlorinated ethanes and dioxins are much better with 0.40 and 0.49 kcal/mol RMSE, respectively. The predictions for the chlorinated biphenyls show a systematic shift of approximately 1 kcal/mol, but the large RMSE of 1.59 kcal/mol mainly arises from two exceptional outliers. Possible reasons for this observation are discussed.

Reliable Quantum Chemical Prediction of the Localized/Delocalized Character of Organic Mixed-Valence Radical Anions. From Continuum Solvent Models to Direct-COSMO-RS.

A recently proposed quantum-chemical protocol for the description of the character of organic mixed-valence (MV) compounds, close from both sides to the localized/delocalized borderline, is evaluated and extended for a series of dinitroaryl radical anions 1-6. A combination of global hybrid functionals with exact-exchange admixtures of 35% (BLYP35) or 42% (BMK) with appropriate solvent modeling allows an essentially quantitative treatment of, for example, structural symmetry-breaking in Robin/Day class II systems, thermal electron transfer (ET) barriers, and intervalence charge-transfer (IV-CT) excitation energies, while covering also the delocalized class III cases. Global hybrid functionals with lower exact-exchange admixtures (e.g., B3LYP, M05, or M06) provide a too delocalized description, while functionals with higher exact-exchange admixtures (M05-2X, M06-2X) provide a too localized one. The B2PLYP double hybrid gives reasonable structures but far too small barriers in class II cases. The CAM-B3LYP range hybrid gives somewhat too high ET barriers and IV-CT energies, while the range hybrids ωB97X and LC-BLYP clearly exhibit too much exact exchange. Continuum solvent models describe the situation well in most aprotic solvents studied. The transition of 1,4-dinitrobenzene anion 1 from a class III behavior in aprotic solvents to a class II behavior in alcohols is not recovered by continuum solvent models. In contrast, it is treated faithfully by the novel direct conductor-like screening model for real solvents (D-COSMO-RS). The D-COSMO-RS approach, the TURBOMOLE implementation of which is reported, also describes accurately the increased ET barriers of class II systems 2 and 3 in alcohols as compared to aprotic solvents and can distinguish at least qualitatively between different aprotic solvents with identical or similar dielectric constants. The dominant role of the solvent environment for the ET character of these MV radical anions is emphasized, as in contrast to some previous computational suggestions essentially all of the present systems have delocalized class III character in the gas phase. The present approach allows accurate estimates from the gas phase to aprotic and protic solvent environments, without the need for explicit ab initio molecular dynamics simulations, and without artificial constraints.