Using Machine Learning to Parse Chemical Mixture Descriptions.

Chemical mixtures have recently come to the attention of open standards and data structures for capturing machine-readable descriptions for informatics uses. At the present time, essentially all transmission of information about mixtures is done using short text descriptions that are readable only by trained scientists, and there are no accessible repositories of marked-up mixture data. We have designed a machine learning tool that can interpret mixture descriptions and upgrade them to the high-level Mixfile format, which can in turn be used to generate Mixtures InChI notation. The interpretation achieves a high success rate and can be used at scale to markup large catalogs and inventories, with some expert checking to catch edge cases. The training data that was accumulated during the project is made openly available, along with previously released mixture editing tools and utilities.

Prediction of pKa Using Machine Learning Methods with Rooted Topological Torsion Fingerprints: Application to Aliphatic Amines.

The acid-base dissociation constant, pKa, is a key parameter to define the ionization state of a compound and directly affects its biopharmaceutical profile. In this study, we developed a novel approach for pKa prediction using rooted topological torsion fingerprints in combination with five machine learning (ML) methods: random forest, partial least squares, extreme gradient boosting, lasso regression, and support vector regression. With a large and diverse set of 14 499 experimental pKa values, pKa models were developed for aliphatic amines. The models demonstrated consistently good prediction statistics and were able to generate accurate prospective predictions as validated with an external test set of 726 pKa values (RMSE 0.45, MAE 0.33, and R2 0.84 by the top model). The factors that may affect prediction accuracy and model applicability were carefully assessed. The results demonstrated that rooted topological torsion fingerprints coupled with ML methods provide a promising approach for developing accurate pKa prediction models.

Capturing mixture composition: an open machine-readable format for representing mixed substances.

We describe a file format that is designed to represent mixtures of compounds in a way that is fully machine readable. This Mixfile format is intended to fill the same role for substances that are composed of multiple components as the venerable Molfile does for specifying individual structures. This much needed datastructure is intended to replace current practices for communicating information about mixtures, which usually relies on human-readable text descriptions, drawing several species within a single molecular diagram, or mutually incompatible ad hoc solutions. We describe an open source software application for editing mixture files, which can also be used as web-ready tools for manipulating the file format. We also present a corpus of mixture examples, which we have extracted from collections of text-based descriptions. Furthermore, we present an early look at the proposed IUPAC Mixtures InChI specification, instances of which can be automatically generated using the Mixfile format as a precursor.

Development of a Cytopathic Effect-Based Phenotypic Screening Assay against Cryptosporidium.

Cryptosporidiosis is a diarrheal disease predominantly caused by Cryptosporidium parvum ( Cp) and Cryptosporidium hominis ( Ch), apicomplexan parasites which infect the intestinal epithelial cells of their human hosts. The only approved drug for cryptosporidiosis is nitazoxanide, which shows limited efficacy in immunocompromised children, the most vulnerable patient population. Thus, new therapeutics and in vitro infection models are urgently needed to address the current unmet medical need. Toward this aim, we have developed novel cytopathic effect (CPE)-based Cp and Ch assays in human colonic tumor (HCT-8) cells and compared them to traditional imaging formats. Further model validation was achieved through screening a collection of FDA-approved drugs and confirming many previously known anti- Cryptosporidium hits as well as identifying a few novel candidates. Collectively, our data reveals this model to be a simple, functional, and homogeneous gain of signal format amenable to high throughput screening, opening new avenues for the discovery of novel anticryptosporidials.

Developing Collaborative QSAR Models Without Sharing Structures.

It is widely understood that QSAR models greatly improve if more data are used. However, irrespective of model quality, once chemical structures diverge too far from the initial data set, the predictive performance of a model degrades quickly. To increase the applicability domain we need to increase the diversity of the training set. This can be achieved by combining data from diverse sources. Public data can be easily included; however, proprietary data may be more difficult to add due to intellectual property concerns. In this contribution, we will present a method for the collaborative development of linear regression models that addresses this problem. The method differs from other past approaches, because data are only shared in an aggregated form. This prohibits access to individual data points and therefore avoids the disclosure of confidential structural information. The final models are equivalent to models that were built with combined data sets.

A Cryptosporidium PI(4)K inhibitor is a drug candidate for cryptosporidiosis.

Diarrhoeal disease is responsible for 8.6% of global child mortality. Recent epidemiological studies found the protozoan parasite Cryptosporidium to be a leading cause of paediatric diarrhoea, with particularly grave impact on infants and immunocompromised individuals. There is neither a vaccine nor an effective treatment. Here we establish a drug discovery process built on scalable phenotypic assays and mouse models that take advantage of transgenic parasites. Screening a library of compounds with anti-parasitic activity, we identify pyrazolopyridines as inhibitors of Cryptosporidium parvum and Cryptosporidium hominis. Oral treatment with the pyrazolopyridine KDU731 results in a potent reduction in intestinal infection of immunocompromised mice. Treatment also leads to rapid resolution of diarrhoea and dehydration in neonatal calves, a clinical model of cryptosporidiosis that closely resembles human infection. Our results suggest that the Cryptosporidium lipid kinase PI(4)K (phosphatidylinositol-4-OH kinase) is a target for pyrazolopyridines and that KDU731 warrants further preclinical evaluation as a drug candidate for the treatment of cryptosporidiosis.

Shared Consensus Machine Learning Models for Predicting Blood Stage Malaria Inhibition.

The development of new antimalarial therapies is essential, and lowering the barrier of entry for the screening and discovery of new lead compound classes can spur drug development at organizations that may not have large compound screening libraries or resources to conduct high-throughput screens. Machine learning models have been long established to be more robust and have a larger domain of applicability with larger training sets. Screens over multiple data sets to find compounds with potential malaria blood stage inhibitory activity have been used to generate multiple Bayesian models. Here we describe a method by which Bayesian quantitative structure-activity relationship models, which contain information on thousands to millions of proprietary compounds, can be shared between collaborators at both for-profit and not-for-profit institutions. This model-sharing paradigm allows for the development of consensus models that have increased predictive power over any single model and yet does not reveal the identity of any compounds in the training sets.

UDP-galactose and acetyl-CoA transporters as Plasmodium multidrug resistance genes.

A molecular understanding of drug resistance mechanisms enables surveillance of the effectiveness of new antimicrobial therapies during development and deployment in the field. We used conventional drug resistance selection as well as a regime of limiting dilution at early stages of drug treatment to probe two antimalarial imidazolopiperazines, KAF156 and GNF179. The latter approach permits the isolation of low-fitness mutants that might otherwise be out-competed during selection. Whole-genome sequencing of 24 independently derived resistant Plasmodium falciparum clones revealed four parasites with mutations in the known cyclic amine resistance locus (pfcarl) and a further 20 with mutations in two previously unreported P. falciparum drug resistance genes, an acetyl-CoA transporter (pfact) and a UDP-galactose transporter (pfugt). Mutations were validated both in vitro by CRISPR editing in P. falciparum and in vivo by evolution of resistant Plasmodium berghei mutants. Both PfACT and PfUGT were localized to the endoplasmic reticulum by fluorescence microscopy. As mutations in pfact and pfugt conveyed resistance against additional unrelated chemical scaffolds, these genes are probably involved in broad mechanisms of antimalarial drug resistance.

Mutations in the Plasmodium falciparum Cyclic Amine Resistance Locus (PfCARL) Confer Multidrug Resistance.

UNLABELLED: Mutations in the Plasmodium falciparum cyclic amine resistance locus (PfCARL) are associated with parasite resistance to the imidazolopiperazines, a potent class of novel antimalarial compounds that display both prophylactic and transmission-blocking activity, in addition to activity against blood-stage parasites. Here, we show that pfcarl encodes a protein, with a predicted molecular weight of 153 kDa, that localizes to the cis-Golgi apparatus of the parasite in both asexual and sexual blood stages. Utilizing clustered regularly interspaced short palindromic repeat (CRISPR)-mediated gene introduction of 5 variants (L830V, S1076N/I, V1103L, and I1139K), we demonstrate that mutations in pfcarl are sufficient to generate resistance against the imidazolopiperazines in both asexual and sexual blood-stage parasites. We further determined that the mutant PfCARL protein confers resistance to several structurally unrelated compounds. These data suggest that PfCARL modulates the levels of small-molecule inhibitors that affect Golgi-related processes, such as protein sorting or membrane trafficking, and is therefore an important mechanism of resistance in malaria parasites. IMPORTANCE: Several previous in vitro evolution studies have implicated the Plasmodium falciparum cyclic amine resistance locus (PfCARL) as a potential target of imidazolopiperazines, potent antimalarial compounds with broad activity against different parasite life cycle stages. Given that the imidazolopiperazines are currently being tested in clinical trials, understanding their mechanism of resistance and the cellular processes involved will allow more effective clinical usage.

Benefit of Retraining pKa Models Studied Using Internally Measured Data.

The ionization state of drugs influences many pharmaceutical properties such as their solubility, permeability, and biological activity. It is therefore important to understand the structure property relationship for the acid-base dissociation constant pKa during the lead optimization process to make better-informed design decisions. Computational approaches, such as implemented in MoKa, can help with this; however, they often predict with too large error especially for proprietary compounds. In this contribution, we look at how retraining helps to greatly improve prediction error. Using a longitudinal study with data measured over 15 years in a drug discovery environment, we assess the impact of model training on prediction accuracy and look at model degradation over time. Using the MoKa software, we will demonstrate that regular retraining is required to address changes in chemical space leading to model degradation over six to nine months.

FOCUS--Development of a Global Communication and Modeling Platform for Applied and Computational Medicinal Chemists.

Communication of data and ideas within a medicinal chemistry project on a global as well as local level is a crucial aspect in the drug design cycle. Over a time frame of eight years, we built and optimized FOCUS, a platform to produce, visualize, and share information on various aspects of a drug discovery project such as cheminformatics, data analysis, structural information, and design. FOCUS is tightly integrated with internal services that involve-among others-data retrieval systems and in-silico models and provides easy access to automated modeling procedures such as pharmacophore searches, R-group analysis, and similarity searches. In addition, an interactive 3D editor was developed to assist users in the generation and docking of close analogues of a known lead. In this paper, we will specifically concentrate on issues we faced during development, deployment, and maintenance of the software and how we continually adapted the software in order to improve usability. We will provide usage examples to highlight the functionality as well as limitations of FOCUS at the various stages of the development process. We aim to make the discussion as independent of the software platform as possible, so that our experiences can be of more general value to the drug discovery community.

Matched molecular pair analysis: significance and the impact of experimental uncertainty.

Matched molecular pair analysis (MMPA) has become a major tool for analyzing large chemistry data sets for promising chemical transformations. However, the dependence of MMPA predictions on data constraints such as the number of pairs involved, experimental uncertainty, source of the experiments, and variability of the true physical effect has not yet been described. In this contribution the statistical basics for judging MMPA are analyzed. We illustrate the connection between overall MMPA statistics and individual pairs with a detailed comparison of average CHEMBL hERG MMPA results versus pairs with extreme transformation effects. Comparing the CHEMBL results to Novartis data, we find that significant transformation effects agree very well if the experimental uncertainty is considered. This indicates that caution must be exercised for predictions from insignificant MMPAs, yet highlights the robustness of statistically validated MMPA and shows that MMPA on public databases can yield results that are very useful for medicinal chemistry.

Deriving static atomic multipoles from the electrostatic potential.

The description of molecular systems using multipolar electrostatics calls for automated methods to fit the necessary parameters. In this paper, we describe an open-source software package that allows fitting atomic multipoles (MTPs) from the ab initio electrostatic potential by adequate atom typing and judicious assignment of the local axis system. By enabling the simultaneous fit of several molecules and/or conformations, the package addresses issues of parameter transferability and lack of sampling for buried atoms. We illustrate the method by studying a series of small alcohol molecules, as well as various conformations of protonated butylamine.

Comparability of mixed IC50 data - a statistical analysis.

The biochemical half maximal inhibitory concentration (IC50) is the most commonly used metric for on-target activity in lead optimization. It is used to guide lead optimization, build large-scale chemogenomics analysis, off-target activity and toxicity models based on public data. However, the use of public biochemical IC50 data is problematic, because they are assay specific and comparable only under certain conditions. For large scale analysis it is not feasible to check each data entry manually and it is very tempting to mix all available IC50 values from public database even if assay information is not reported. As previously reported for Ki database analysis, we first analyzed the types of errors, the redundancy and the variability that can be found in ChEMBL IC50 database. For assessing the variability of IC50 data independently measured in two different labs at least ten IC50 data for identical protein-ligand systems against the same target were searched in ChEMBL. As a not sufficient number of cases of this type are available, the variability of IC50 data was assessed by comparing all pairs of independent IC50 measurements on identical protein-ligand systems. The standard deviation of IC50 data is only 25% larger than the standard deviation of Ki data, suggesting that mixing IC50 data from different assays, even not knowing assay conditions details, only adds a moderate amount of noise to the overall data. The standard deviation of public ChEMBL IC50 data, as expected, resulted greater than the standard deviation of in-house intra-laboratory/inter-day IC50 data. Augmenting mixed public IC50 data by public Ki data does not deteriorate the quality of the mixed IC50 data, if the Ki is corrected by an offset. For a broad dataset such as ChEMBL database a Ki- IC50 conversion factor of 2 was found to be the most reasonable.

Multipole-Based Force Fields from ab Initio Interaction Energies and the Need for Jointly Refitting All Intermolecular Parameters.

Distributed atomic multipole (MTP) moments promise significant improvements over point charges (PCs) in molecular force fields, as they (a) more realistically reproduce the ab initio electrostatic potential (ESP) and (b) allow to capture anisotropic atomic properties such as lone pairs, conjugated systems, and σ holes. The present work focuses on the question of whether multipolar electrostatics instead of PCs in standard force fields leads to quantitative improvements over point charges in reproducing intermolecular interactions. To this end, the interaction energies of two model systems, benzonitrile (BZN) and formamide (FAM) homodimers, are characterized over a wide range of dimer conformations. It is found that although with MTPs the monomer ab initio ESP can be captured better by about an order of magnitude compared to point charges (PCs), this does not directly translate into better describing ab initio interaction energies compared to PCs. Neither ESP-fitted MTPs nor refitted Lennard-Jones (LJ) parameters alone demonstrate a clear superiority of atomic MTPs. We show that only if both electrostatic and LJ parameters are jointly optimized in standard, nonpolarizable force fields, atomic are MTPs clearly beneficial for reproducing ab initio dimerization energies. After an exhaustive exponent scan, we find that for both BZN and FAM, atomic MTPs and a 9-6 LJ potential can reproduce ab initio interaction energies with ∼30% (RMSD 0.13 vs 0.18 kcal/mol) less error than point charges (PCs) and a 12-6 LJ potential. We also find that the improvement due to using MTPs with a 9-6 LJ potential is considerably more pronounced than with a 12-6 LJ potential (≈ 10%; RMSD 0.19 versus 0.21 kcal/mol).

Development of isoform selective PI3-kinase inhibitors as pharmacological tools for elucidating the PI3K pathway.

Using a parallel synthesis approach to target a non-conserved region of the PI3K catalytic domain a pan-PI3K inhibitor 1 was elaborated to provide alpha, delta and gamma isoform selective Class I PI3K inhibitors 21, 24, 26 and 27. The compounds had good cellular activity and were selective against protein kinases and other members of the PI3K superfamily including mTOR and DNA-PK.

The experimental uncertainty of heterogeneous public K(i) data.

The maximum achievable accuracy of in silico models depends on the quality of the experimental data. Consequently, experimental uncertainty defines a natural upper limit to the predictive performance possible. Models that yield errors smaller than the experimental uncertainty are necessarily overtrained. A reliable estimate of the experimental uncertainty is therefore of high importance to all originators and users of in silico models. The data deposited in ChEMBL was analyzed for reproducibility, i.e., the experimental uncertainty of independent measurements. Careful filtering of the data was required because ChEMBL contains unit-transcription errors, undifferentiated stereoisomers, and repeated citations of single measurements (90% of all pairs). The experimental uncertainty is estimated to yield a mean error of 0.44 pK(i) units, a standard deviation of 0.54 pK(i) units, and a median error of 0.34 pK(i) units. The maximum possible squared Pearson correlation coefficient (R(2)) on large data sets is estimated to be 0.81.

Atomic multipoles: electrostatic potential fit, local reference axis systems, and conformational dependence.

Currently, all standard force fields for biomolecular simulations use point charges to model intermolecular electrostatic interactions. This is a fast and simple approach but has deficiencies when the electrostatic potential (ESP) is compared to that from ab initio methods. Here, we show how atomic multipoles can be rigorously implemented into common biomolecular force fields. For this, a comprehensive set of local reference axis systems is introduced, which represents a universal solution for treating atom-centered multipoles for all small organic molecules and proteins. Furthermore, we introduce a new method for fitting atomic multipole moments to the quantum mechanically derived ESP. This methods yields a 50-90% error reduction compared to both point charges fit to the ESP and multipoles directly calculated from the ab initio electron density. It is shown that it is necessary to directly fit the multipole moments of conformational ensembles to the ESP. Ignoring the conformational dependence or averaging over parameters from different conformations dramatically deteriorates the results obtained with atomic multipole moments, rendering multipoles worse than partial charges.

Design and synthesis of a library of chemokine antagonists.

A library of chemokine antagonists has been synthesized using a combination of solid and solution-phase chemistry. Structures of known chemokine antagonists were used to produce a pharmacophore which served to guide monomer selection. Several combinations of monomers have resulted in providing novel chemokine antagonists which in some cases display dual chemokine receptor antagonism.

Three descriptor model sets a high standard for the CSAR-NRC HiQ benchmark.

Here we report the results we obtained with a proteochemometric approach for predicting ligand binding free energies of the CSAR-NRC HiQ benchmark data set. Using distance-dependent atom-type pair descriptors in a bagged stepwise multiple-linear regression (MLR) model with subsequent complexity reduction we were able to identify three descriptors that can be used to build a very robust regression model for the CSAR-NRC HiQ data set. The model has an R(2)(cv) of 0.55, a MUE(cv) of 1.19, and an RMSE(cv) of 1.49 on the out-of-bag test set. The descriptors selected are the count of protein atoms in a shell between 4.5 Å and 6 Å around each heavy ligand atom excluding oxygen and phosphorus, the count of sulfur atoms in the vicinity of tryptophan, and the count of aliphatic ligand hydroxy hydrogens. The first two descriptors have a positive sign indicating that they contribute favorably to the binding energy, whereas the count of hydroxy hydrogens contributes unfavorably to the binding free energy observed. The fact that such a simple model can be so effective raises a couple of questions that are addressed in the article.