Mechanistic Insights into Passive Membrane Permeability of Drug-like Molecules from a Weighted Ensemble of Trajectories.

Passive permeability of a drug-like molecule is a critical property assayed early in a drug discovery campaign that informs a medicinal chemist how well a compound can traverse biological membranes, such as gastrointestinal epithelial or restrictive organ barriers, so it can perform a specific therapeutic function. However, the challenge that remains is the development of a method, experimental or computational, which can both determine the permeation rate and provide mechanistic insights into the transport process to help with the rational design of any given molecule. Typically, one of the following three methods are used to measure the membrane permeability: (1) experimental permeation assays acting on either artificial or natural membranes; (2) quantitative structure-permeability relationship models that rely on experimental values of permeability or related pharmacokinetic properties of a range of molecules to infer those for new molecules; and (3) estimation of permeability from the Smoluchowski equation, where free energy and diffusion profiles along the membrane normal are taken as input from large-scale molecular dynamics simulations. While all these methods provide estimates of permeation coefficients, they provide very little information for guiding rational drug design. In this study, we employ a highly parallelizable weighted ensemble (WE) path sampling strategy, empowered by cloud computing techniques, to generate unbiased permeation pathways and permeability coefficients for a set of drug-like molecules across a neat 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylcholine membrane bilayer. Our WE method predicts permeability coefficients that compare well to experimental values from an MDCK-LE cell line and PAMPA assays for a set of drug-like amines of varying size, shape, and flexibility. Our method also yields a series of continuous permeation pathways weighted and ranked by their associated probabilities. Taken together, the ensemble of reactive permeation pathways, along with the estimate of the permeability coefficient, provides a clearer picture of the microscopic underpinnings of small-molecule membrane permeation.

SAMPL6 challenge results from [Formula: see text] predictions based on a general Gaussian process model.

A variety of fields would benefit from accurate [Formula: see text] predictions, especially drug design due to the effect a change in ionization state can have on a molecule's physiochemical properties. Participants in the recent SAMPL6 blind challenge were asked to submit predictions for microscopic and macroscopic [Formula: see text]s of 24 drug like small molecules. We recently built a general model for predicting [Formula: see text]s using a Gaussian process regression trained using physical and chemical features of each ionizable group. Our pipeline takes a molecular graph and uses the OpenEye Toolkits to calculate features describing the removal of a proton. These features are fed into a Scikit-learn Gaussian process to predict microscopic [Formula: see text]s which are then used to analytically determine macroscopic [Formula: see text]s. Our Gaussian process is trained on a set of 2700 macroscopic [Formula: see text]s from monoprotic and select diprotic molecules. Here, we share our results for microscopic and macroscopic predictions in the SAMPL6 challenge. Overall, we ranked in the middle of the pack compared to other participants, but our fairly good agreement with experiment is still promising considering the challenge molecules are chemically diverse and often polyprotic while our training set is predominately monoprotic. Of particular importance to us when building this model was to include an uncertainty estimate based on the chemistry of the molecule that would reflect the likely accuracy of our prediction. Our model reports large uncertainties for the molecules that appear to have chemistry outside our domain of applicability, along with good agreement in quantile-quantile plots, indicating it can predict its own accuracy. The challenge highlighted a variety of means to improve our model, including adding more polyprotic molecules to our training set and more carefully considering what functional groups we do or do not identify as ionizable.

Efficient calculation of SAMPL4 hydration free energies using OMEGA, SZYBKI, QUACPAC, and Zap TK.

Several submissions for the SAMPL4 hydration free energy set were calculated using OpenEye tools, including many that were among the top performing submissions. All of our best submissions used AM1BCC charges and Poisson-Boltzmann solvation. Three submissions used a single conformer for calculating the hydration free energy and all performed very well with mean unsigned errors ranging from 0.94 to 1.08 kcal/mol. These calculations were very fast, only requiring 0.5-2.0 s per molecule. We observed that our two single-conformer methodologies have different types of failure cases and that these differences could be exploited for determining when the methods are likely to have substantial errors.

The SAMPL2 blind prediction challenge: introduction and overview.

The interactions between a molecule and the aqueous environment underpin any process that occurs in solution, from simple chemical reactions to protein-ligand binding to protein aggregation. Fundamental measures of the interaction between molecule and aqueous phase, such as the transfer energy between gas phase and water or the energetic difference between two tautomers of a molecule in solution, remain nontrivial to predict accurately using current computational methods. SAMPL2 represents the third annual blind prediction of transfer energies, and the first time tautomer ratios were included in the challenge. Over 60 sets of predictions were submitted, and each participant also attempted to estimate the error in their predictions, a task that proved difficult for most. The results of this blind assessment of the state of the field for transfer energy and tautomer ratio prediction both indicate where the field is performing well and point out flaws in current methods.

Analysis of SM8 and Zap TK calculations and their geometric sensitivity.

A prospective study of aqueous solvation energies was done using the SM8 and Zap TK models for a variety of geometries. CM4M charges calculated with M06 and M06-2X were found to yield similar results for the SM8 model. Zap TK calculations were primarily done with AM1BCC charges but limited attempts to use charges derived from DFT showed promise. The OMEGA application quickly generated conformations that performed well with both solvation models, while the use of computationally expensive DFT optimized geometries yielded increased RMSE and MSE. It is shown that increasing levels of gas phase geometry optimization yield increasingly unfavorable solvation energy for single conformer models.

Conformer generation with OMEGA: algorithm and validation using high quality structures from the Protein Databank and Cambridge Structural Database.

Here, we present the algorithm and validation for OMEGA, a systematic, knowledge-based conformer generator. The algorithm consists of three phases: assembly of an initial 3D structure from a library of fragments; exhaustive enumeration of all rotatable torsions using values drawn from a knowledge-based list of angles, thereby generating a large set of conformations; and sampling of this set by geometric and energy criteria. Validation of conformer generators like OMEGA has often been undertaken by comparing computed conformer sets to experimental molecular conformations from crystallography, usually from the Protein Databank (PDB). Such an approach is fraught with difficulty due to the systematic problems with small molecule structures in the PDB. Methods are presented to identify a diverse set of small molecule structures from cocomplexes in the PDB that has maximal reliability. A challenging set of 197 high quality, carefully selected ligand structures from well-solved models was obtained using these methods. This set will provide a sound basis for comparison and validation of conformer generators in the future. Validation results from this set are compared to the results using structures of a set of druglike molecules extracted from the Cambridge Structural Database (CSD). OMEGA is found to perform very well in reproducing the crystallographic conformations from both these data sets using two complementary metrics of success.

In vitro and in vivo activity of novel small-molecule inhibitors targeting the pleckstrin homology domain of protein kinase B/AKT.

The phosphatidylinositol 3-kinase/AKT signaling pathway plays a critical role in activating survival and antiapoptotic pathways within cancer cells. Several studies have shown that this pathway is constitutively activated in many different cancer types. The goal of this study was to discover novel compounds that bind to the pleckstrin homology (PH) domain of AKT, thereby inhibiting AKT activation. Using proprietary docking software, 22 potential PH domain inhibitors were identified. Surface plasmon resonance spectroscopy was used to measure the binding of the compounds to the expressed PH domain of AKT followed by an in vitro activity screen in Panc-1 and MiaPaCa-2 pancreatic cancer cell lines. We identified a novel chemical scaffold in several of the compounds that binds selectively to the PH domain of AKT, inducing a decrease in AKT activation and causing apoptosis at low micromolar concentrations. Structural modifications of the scaffold led to compounds with enhanced inhibitory activity in cells. One compound, 4-dodecyl-N-(1,3,4-thiadiazol-2-yl)benzenesulfonamide, inhibited AKT and its downstream targets in cells as well as in pancreatic cancer cell xenografts in immunocompromised mice; it also exhibited good antitumor activity. In summary, a pharmacophore for PH domain inhibitors targeting AKT function was developed. Computer-aided modeling, synthesis, and testing produced novel AKT PH domain inhibitors that exhibit promising preclinical properties.

How to do an evaluation: pitfalls and traps.

The recent literature is replete with papers evaluating computational tools (often those operating on 3D structures) for their performance in a certain set of tasks. Most commonly these papers compare a number of docking tools for their performance in cognate re-docking (pose prediction) and/or virtual screening. Related papers have been published on ligand-based tools: pose prediction by conformer generators and virtual screening using a variety of ligand-based approaches. The reliability of these comparisons is critically affected by a number of factors usually ignored by the authors, including bias in the datasets used in virtual screening, the metrics used to assess performance in virtual screening and pose prediction and errors in crystal structures used.

Comparison of shape-matching and docking as virtual screening tools.

Ligand docking is a widely used approach in virtual screening. In recent years a large number of publications have appeared in which docking tools are compared and evaluated for their effectiveness in virtual screening against a wide variety of protein targets. These studies have shown that the effectiveness of docking in virtual screening is highly variable due to a large number of possible confounding factors. Another class of method that has shown promise in virtual screening is the shape-based, ligand-centric approach. Several direct comparisons of docking with the shape-based tool ROCS have been conducted using data sets from some of these recent docking publications. The results show that a shape-based, ligand-centric approach is more consistent than, and often superior to, the protein-centric approach taken by docking.

A novel mechanism for inhibition of HIV-1 reverse transcriptase.

The human immunodeficiency virus (HIV) epidemic is an important medical problem. Although combination drug regimens have produced dramatic decreases in viral load, current therapies do not provide a cure for HIV infection. We have used structure-based design and combinatorial medicinal chemistry to identify potent and selective HIV-1 reverse transcriptase (RT) inhibitors that may work by a mechanism distinct from that of current HIV drugs. The most potent of these compounds (compound 4, 2-naphthalenesulfonic acid, 4-hydroxy-7-[[[[5-hydroxy-6-[(4-cinnamylphenyl)azo]-7-sulfo-2-naphthalenyl]amino]carbonyl]amino]-3-[(4-cinnamylphenyl)azo], disodium salt) has an IC(50) of 90 nM for inhibition of polymerase chain extension, a K(d) of 40 nM for inhibition of DNA-RT binding, and an IC(50) of 25-100 nM for inhibition of RNaseH cleavage. The parent compound (1) was as effective against 10 nucleoside and non-nucleoside resistant HIV-1 RT mutants as it was against the wild-type enzyme. Compound 4 inhibited HIV-1 RT and murine leukemia virus (MLV) RT, but it did not inhibit T(4) DNA polymerase, T(7) DNA polymerase, or the Klenow fragment at concentrations up to 200 nM. Finally, compound 4 protected cells from HIV-1 infection at a concentration more than 40 times lower than the concentration at which it caused cellular toxicity.