Mechanistic Reactivity Descriptors for the Prediction of Ames Mutagenicity of Primary Aromatic Amines.

Pharmaceutical products are often synthesized by the use of reactive starting materials and intermediates. These can, either as impurities or through metabolic activation, bind to the DNA. Primary aromatic amines belong to the critical classes that are considered potentially mutagenic in the Ames test, so there is a great need for good prediction models for risk assessment. How primary aromatic amines exert their mutagenic potential can be rationalized by the widely accepted nitrenium ion hypothesis of covalent binding to the DNA of reactive electrophiles formed out of the aromatic amines. Since the reactive chemical species is different in chemical structure from the actual compound, it is difficult to achieve good predictions via classical descriptor or fingerprint-based machine learning. In this approach, we use a combination of different molecular and atomic descriptors that is able to describe different mechanistic aspects of the metabolic transformation leading from the primary aromatic amine to the reactive metabolite that binds to the DNA. Applied to a test set, the combination shows significantly better performance than models that only use one of these descriptors and complemented the general internal Ames mutagenicity prediction model at Bayer.

Modeling Physico-Chemical ADMET Endpoints with Multitask Graph Convolutional Networks.

Simple physico-chemical properties, like logD, solubility, or melting point, can reveal a great deal about how a compound under development might later behave. These data are typically measured for most compounds in drug discovery projects in a medium throughput fashion. Collecting and assembling all the Bayer in-house data related to these properties allowed us to apply powerful machine learning techniques to predict the outcome of those assays for new compounds. In this paper, we report our finding that, especially for predicting physicochemical ADMET endpoints, a multitask graph convolutional approach appears a highly competitive choice. For seven endpoints of interest, we compared the performance of that approach to fully connected neural networks and different single task models. The new model shows increased predictive performance compared to previous modeling methods and will allow early prioritization of compounds even before they are synthesized. In addition, our model follows the generalized solubility equation without being explicitly trained under this constraint.

Extending (Q)SARs to incorporate proprietary knowledge for regulatory purposes: A case study using aromatic amine mutagenicity.

Statistical-based and expert rule-based models built using public domain mutagenicity knowledge and data are routinely used for computational (Q)SAR assessments of pharmaceutical impurities in line with the approach recommended in the ICH M7 guideline. Knowledge from proprietary corporate mutagenicity databases could be used to increase the predictive performance for selected chemical classes as well as expand the applicability domain of these (Q)SAR models. This paper outlines a mechanism for sharing knowledge without the release of proprietary data. Primary aromatic amine mutagenicity was selected as a case study because this chemical class is often encountered in pharmaceutical impurity analysis and mutagenicity of aromatic amines is currently difficult to predict. As part of this analysis, a series of aromatic amine substructures were defined and the number of mutagenic and non-mutagenic examples for each chemical substructure calculated across a series of public and proprietary mutagenicity databases. This information was pooled across all sources to identify structural classes that activate or deactivate aromatic amine mutagenicity. This structure activity knowledge, in combination with newly released primary aromatic amine data, was incorporated into Leadscope's expert rule-based and statistical-based (Q)SAR models where increased predictive performance was demonstrated.

Discovery of YAP1/TAZ pathway inhibitors through phenotypic screening with potent anti-tumor activity via blockade of Rho-GTPase signaling.

This study describes the identification and target deconvolution of small molecule inhibitors of oncogenic Yes-associated protein (YAP1)/TAZ activity with potent anti-tumor activity in vivo. A high-throughput screen (HTS) of 3.8 million compounds was conducted using a cellular YAP1/TAZ reporter assay. Target deconvolution studies identified the geranylgeranyltransferase-I (GGTase-I) complex as the direct target of YAP1/TAZ pathway inhibitors. The small molecule inhibitors block the activation of Rho-GTPases, leading to subsequent inactivation of YAP1/TAZ and inhibition of cancer cell proliferation in vitro. Multi-parameter optimization resulted in BAY-593, an in vivo probe with favorable PK properties, which demonstrated anti-tumor activity and blockade of YAP1/TAZ signaling in vivo.

Trends in Molecular Properties, Bioavailability, and Permeability across the Bayer Compound Collection.

For oral drugs, medicinal chemists aim to design compounds with high oral bioavailability, of which permeability is a key determinant. Taking advantage of >2000 compounds tested in rat bioavailability studies and >20,000 compounds tested in Caco2 assays at Bayer, we have examined the molecular properties governing bioavailability and permeability. In addition to classical parameters such as logD and molecular weight, we also investigated the relationship between calculated pKa and permeability. We find that neutral compounds retain permeability up to a molecular weight limit of 700, while stronger acids and bases are restricted to weights of 400-500. We also investigate trends for common properties such as hydrogen bond donors and acceptors, polar surface area, aromatic ring count, and rotatable bonds, including compounds which exceed Lipinski's rule of five (Ro5). These property-structure relationships are combined to provide design guidelines for bioavailable drugs in both traditional and "beyond rule of 5" (bRo5) chemical space.

Machine Learning Applied to the Modeling of Pharmacological and ADMET Endpoints.

The well-known concept of quantitative structure-activity relationships (QSAR) has been gaining significant interest in the recent years. Data, descriptors, and algorithms are the main pillars to build useful models that support more efficient drug discovery processes with in silico methods. Significant advances in all three areas are the reason for the regained interest in these models. In this book chapter we review various machine learning (ML) approaches that make use of measured in vitro/in vivo data of many compounds. We put these in context with other digital drug discovery methods and present some application examples.

The potent AMPK inhibitor BAY-3827 shows strong efficacy in androgen-dependent prostate cancer models.

PURPOSE: 5' adenosine monophosphate-activated kinase (AMPK) is an essential regulator of cellular energy homeostasis and has been associated with different pathologies, including cancer. Precisely defining the biological role of AMPK necessitates the availability of a potent and selective inhibitor. METHODS: High-throughput screening and chemical optimization were performed to identify a novel AMPK inhibitor. Cell proliferation and mechanistic assays, as well as gene expression analysis and chromatin immunoprecipitation were used to investigate the cellular impact as well as the crosstalk between lipid metabolism and androgen signaling in prostate cancer models. Also, fatty acid turnover was determined by examining lipid droplet formation. RESULTS: We identified BAY-3827 as a novel and potent AMPK inhibitor with additional activity against ribosomal 6 kinase (RSK) family members. It displays strong anti-proliferative effects in androgen-dependent prostate cancer cell lines. Analysis of genes involved in AMPK signaling revealed that the expression of those encoding 3-hydroxy-3-methyl-glutaryl-coenzyme A reductase (HMGCR), fatty acid synthase (FASN) and 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 2 (PFKFB2), all of which are involved in lipid metabolism, was strongly upregulated by androgen in responsive models. Chromatin immunoprecipitation DNA-sequencing (ChIP-seq) analysis identified several androgen receptor (AR) binding peaks in the HMGCR and PFKFB2 genes. BAY-3827 strongly down-regulated the expression of lipase E (LIPE), cAMP-dependent protein kinase type II-beta regulatory subunit (PRKAR2B) and serine-threonine kinase AKT3 in responsive prostate cancer cell lines. Also, the expression of members of the carnitine palmitoyl-transferase 1 (CPT1) family was inhibited by BAY-3827, and this was paralleled by impaired lipid flux. CONCLUSIONS: The availability of the potent inhibitor BAY-3827 will contribute to a better understanding of the role of AMPK signaling in cancer, especially in prostate cancer.

Bayer's in silico ADMET platform: a journey of machine learning over the past two decades.

Over the past two decades, an in silico absorption, distribution, metabolism, and excretion (ADMET) platform has been created at Bayer Pharma with the goal to generate models for a variety of pharmacokinetic and physicochemical endpoints in early drug discovery. These tools are accessible to all scientists within the company and can be a useful in assisting with the selection and design of novel leads, as well as the process of lead optimization. Here. we discuss the development of machine-learning (ML) approaches with special emphasis on data, descriptors, and algorithms. We show that high company internal data quality and tailored descriptors, as well as a thorough understanding of the experimental endpoints, are essential to the utility of our models. We discuss the recent impact of deep neural networks and show selected application examples.