Exploration of Ultralarge Compound Collections for Drug Discovery.

Designing new medicines more cheaply and quickly is tightly linked to the quest of exploring chemical space more widely and efficiently. Chemical space is monumentally large, but recent advances in computer software and hardware have enabled researchers to navigate virtual chemical spaces containing billions of chemical structures. This review specifically concerns collections of many millions or even billions of enumerated chemical structures as well as even larger chemical spaces that are not fully enumerated. We present examples of chemical libraries and spaces and the means used to construct them, and we discuss new technologies for searching huge libraries and for searching combinatorially in chemical space. We also cover space navigation techniques and consider new approaches to de novo drug design and the impact of the "autonomous laboratory" on synthesis of designed compounds. Finally, we summarize some other challenges and opportunities for the future.

Context Aware Data-Driven Retrosynthetic Analysis.

Modern drug discovery is an iterative process relying on hypothesis generation through exploitation of available data and hypothesis testing that produces informative results necessary for subsequent rounds of exploration. In this setting, hypothesis generation consists of designing chemical structures likely to meet the pharmaceutically relevant objectives of the discovery project pursued while hypothesis testing involves the compound synthesis and biological assays to query the hypothesis. While much attention has been placed on effective compound design, it is often the case that hypothesis generation efforts lead to novel chemical structure designs with no established chemical synthesis route. We introduce a chemical context aware data-driven method built upon millions of available reactions, with attractive run-time characteristics, to recommend synthetic routes matching a precedent-derived template. Coupled with modern automated synthesis platforms and available building block collections, the method enables drug discovery researchers to identify easy to interpret and implement routes for target compounds. Results of this in-house computer-aided synthesis platform termed ChemoPrint are presented here demonstrating how such tools can bridge chemical synthesis knowledge with synthetic resources and facilitate hypothesis testing, thereby reducing the time required to complete an idea-to-data drug discovery cycle.

Current and Future Roles of Artificial Intelligence in Medicinal Chemistry Synthesis.

Artificial intelligence and machine learning have demonstrated their potential role in predictive chemistry and synthetic planning of small molecules; there are at least a few reports of companies employing in silico synthetic planning into their overall approach to accessing target molecules. A data-driven synthesis planning program is one component being developed and evaluated by the Machine Learning for Pharmaceutical Discovery and Synthesis (MLPDS) consortium, comprising MIT and 13 chemical and pharmaceutical company members. Together, we wrote this perspective to share how we think predictive models can be integrated into medicinal chemistry synthesis workflows, how they are currently used within MLPDS member companies, and the outlook for this field.

Navigating the DNA encoded libraries chemical space.

DNA-encoded library (DEL) technology is a novel ligand identification strategy that allows the synthesis and screening of unprecedented chemical diversity more efficiently than conventional methods. However, no reports have been published to systematically study how to increase the diversity and improve the molecular property space that can be covered with DEL. This report describes the development and application of eDESIGNER, an algorithm that comprehensively generates all possible library designs, enumerates and profiles samples from each library and evaluates them to select the libraries to be synthesized. This tool utilizes suitable on-DNA chemistries and available building blocks to design and identify libraries with a pre-defined molecular weight distribution and maximal diversity compared with compound collections from other sources.

Evaluating the Advantages of Using 3D-Enriched Fragments for Targeting BET Bromodomains.

Fragment-based ligand discovery has been successful in targeting diverse proteins. Despite drug-like molecules having more 3D character, traditional fragment libraries are largely composed of flat, aromatic fragments. The use of 3D-enriched fragments for enhancing library diversity is underexplored especially against protein-protein interactions. Here, we evaluate using 3D-enriched fragments against bromodomains. Bromodomains are highly ligandable, but selectivity remains challenging, particularly for bromodomain and extraterminal (BET) family bromodomains. We screened a 3D-enriched fragment library against BRD4(D1) via 1H CPMG NMR with a protein-observed 19F NMR secondary assay. The screen led to 29% of the hits that are selective over two related bromodomains, BRDT(D1) and BPTF, and the identification of underrepresented chemical bromodomain inhibitor scaffolds. Initial structure-activity relationship studies guided by X-ray crystallography led to a ligand-efficient thiazepane, with good selectivity and affinity for BET bromodomains. These results suggest that the incorporation of 3D-enriched fragments to increase library diversity can benefit bromodomain screening.

Idea2Data: Toward a New Paradigm for Drug Discovery.

Increasing the success rate and throughput of drug discovery will require efficiency improvements throughout the process that is currently used in the pharmaceutical community, including the crucial step of identifying hit compounds to act as drivers for subsequent optimization. Hit identification can be carried out through large compound collection screening and often involves the generation and testing of many hypotheses based on available knowledge. In practice, hypothesis generation can involve the selection of promising chemical structures from compound collections using predictive models built from previous screening/assay results. Available physical collections, typically used during hit identification, are of the order of 106 compounds but represent only a small fraction of the small molecule drug-like chemical space. In an effort to survey a larger portion of chemical space and eliminate inefficiencies during hit identification, we introduce a new process, termed Idea2Data (I2D) that tightly integrates computational and experimental components of the drug discovery process. I2D provides the ability to connect a vast virtual collection of compounds readily synthesizable on automated synthesis systems with computational predictive models for the identification of promising structures. This new paradigm enables researchers to process billions of virtual molecules and select structures that can be prepared on automated systems and made available for biological testing, allowing for timely hypothesis testing and follow-up. Since its introduction, I2D has positively impacted several portfolio efforts through identification of new chemical scaffolds and functionalization of existing scaffolds. In this Innovations paper, we describe the I2D process and present an application for the discovery of new ULK inhibitors.

A retrosynthetic analysis algorithm implementation.

The need for synthetic route design arises frequently in discovery-oriented chemistry organizations. While traditionally finding solutions to this problem has been the domain of human experts, several computational approaches, aided by the algorithmic advances and the availability of large reaction collections, have recently been reported. Herein we present our own implementation of a retrosynthetic analysis method and demonstrate its capabilities in an attempt to identify synthetic routes for a collection of approved drugs. Our results indicate that the method, leveraging on reaction transformation rules learned from a large patent reaction dataset, can identify multiple theoretically feasible synthetic routes and, thus, support research chemist everyday efforts.

Exploring Tunable Hyperparameters for Deep Neural Networks with Industrial ADME Data Sets.

Deep learning has drawn significant attention in different areas including drug discovery. It has been proposed that it could outperform other machine learning algorithms, especially with big data sets. In the field of pharmaceutical industry, machine learning models are built to understand quantitative structure-activity relationships (QSARs) and predict molecular activities, including absorption, distribution, metabolism, and excretion (ADME) properties, using only molecular structures. Previous reports have demonstrated the advantages of using deep neural networks (DNNs) for QSAR modeling. One of the challenges while building DNN models is identifying the hyperparameters that lead to better generalization of the models. In this study, we investigated several tunable hyperparameters of deep neural network models on 24 industrial ADME data sets. We analyzed the sensitivity and influence of five different hyperparameters including the learning rate, weight decay for L2 regularization, dropout rate, activation function, and the use of batch normalization. This paper focuses on strategies and practices for DNN model building. Further, the optimized model for each data set was built and compared with the benchmark models used in production. Based on our benchmarking results, we propose several practices for building DNN QSAR models.

Molecular property prediction: recent trends in the era of artificial intelligence.

Artificial intelligence (AI) has become a powerful tool in many fields, including drug discovery. Among various AI applications, molecular property prediction can have more significant immediate impact to the drug discovery process since most algorithms and methods use predicted properties to evaluate, select, and generate molecules. Herein, we provide a brief review of the state-of-art molecular property prediction methodologies and discuss examples reported recently. We highlight key techniques that have been applied to molecular property prediction such as learned representation, multi-task learning, transfer learning, and federated learning. We also point out some critical but less discussed issues such as data set quality, benchmark, model performance evaluation, and prediction confidence quantification.

The Proximal Lilly Collection: Mapping, Exploring and Exploiting Feasible Chemical Space.

Venturing into the immensity of the small molecule universe to identify novel chemical structure is a much discussed objective of many methods proposed by the chemoinformatics community. To this end, numerous approaches using techniques from the fields of computational de novo design, virtual screening and reaction informatics, among others, have been proposed. Although in principle this objective is commendable, in practice there are several obstacles to useful exploitation of the chemical space. Prime among them are the sheer number of theoretically feasible compounds and the practical concern regarding the synthesizability of the chemical structures conceived using in silico methods. We present the Proximal Lilly Collection initiative implemented at Eli Lilly and Co. with the aims to (i) define the chemical space of small, drug-like compounds that could be synthesized using in-house resources and (ii) facilitate access to compounds in this large space for the purposes of ongoing drug discovery efforts. The implementation of PLC relies on coupling access to available synthetic knowledge and resources with chemo/reaction informatics techniques and tools developed for this purpose. We describe in detail the computational framework supporting this initiative and elaborate on the characteristics of the PLC virtual collection of compounds. As an example of the opportunities provided to drug discovery researchers by easy access to a large, realistically feasible virtual collection such as the PLC, we describe a recent application of the technology that led to the discovery of selective kinase inhibitors.

Discovery of selective RIO2 kinase small molecule ligand.

We report the discovery and initial optimization of diphenpyramide and several of its analogs as hRIO2 kinase ligands. One of these analogs is the most selective hRIO2 ligand reported to date. Diphenpyramide is a Cyclooxygenase 1 and 2 inhibitor that was used as an anti-inflammatory agent. The RIO2 kinase affinity of diphenpyramide was discovered by serendipity while profiling of 13 marketed drugs on a large 456 kinase assay panel. The inhibition values also suggested a relative selectivity of diphenpyramide for RIO2 against the other kinases in the panel. Subsequently three available and eight newly synthesized analogs were assayed, one of which showed a 10 fold increased hRIO2 binding affinity. Additionally, this compound shows significantly better selectivity over assayed kinases, when compared to currently known RIO2 inhibitors. As RIO2 is involved in the biosynthesis of the ribosome and cell cycle regulation, our selective ligand may be useful for the delineation of the biological role of this kinase. This article is part of a Special Issue entitled: Inhibitors of Protein Kinases.

LiSIs: An Online Scientific Workflow System for Virtual Screening.

Modern methods of drug discovery and development in recent years make a wide use of computational algorithms. These methods utilise Virtual Screening (VS), which is the computational counterpart of experimental screening. In this manner the in silico models and tools initial replace the wet lab methods saving time and resources. This paper presents the overall design and implementation of a web based scientific workflow system for virtual screening called, the Life Sciences Informatics (LiSIs) platform. The LiSIs platform consists of the following layers: the input layer covering the data file input; the pre-processing layer covering the descriptors calculation, and the docking preparation components; the processing layer covering the attribute filtering, compound similarity, substructure matching, docking prediction, predictive modelling and molecular clustering; post-processing layer covering the output reformatting and binary file merging components; output layer covering the storage component. The potential of LiSIs platform has been demonstrated through two case studies designed to illustrate the preparation of tools for the identification of promising chemical structures. The first case study involved the development of a Quantitative Structure Activity Relationship (QSAR) model on a literature dataset while the second case study implemented a docking-based virtual screening experiment. Our results show that VS workflows utilizing docking, predictive models and other in silico tools as implemented in the LiSIs platform can identify compounds in line with expert expectations. We anticipate that the deployment of LiSIs, as currently implemented and available for use, can enable drug discovery researchers to more easily use state of the art computational techniques in their search for promising chemical compounds. The LiSIs platform is freely accessible (i) under the GRANATUM platform at: http://www.granatum.org and (ii) directly at: http://lisis.cs.ucy.ac.cy.

Multi-objective optimization methods in drug design.

Drug discovery is a challenging multi-objective problem where numerous pharmaceutically important objectives need to be adequately satisfied for a solution to be found. The problem is characterized by vast, complex solution spaces further perplexed by the presence of conflicting objectives. Multi-objective optimization methods, designed specifically to address such problems, have been introduced to the drug discovery field over a decade ago and have steadily gained in acceptance ever since. This paper reviews the latest multi-objective methods and applications reported in the literature, specifically in quantitative structure­activity modeling, docking, de novo design and library design. Further, the paper reports on related developments in drug discovery research and advances in the multi-objective optimization field.

Selectivity data: assessment, predictions, concordance, and implications.

Could high-quality in silico predictions in drug discovery eventually replace part or most of experimental testing? To evaluate the agreement of selectivity data from different experimental or predictive sources, we introduce the new metric concordance minimum significant ratio (cMSR). Empowered by cMSR, we find the overall level of agreement between predicted and experimental data to be comparable to that found between experimental results from different sources. However, for molecules that are either highly selective or potent, the concordance between different experimental sources is significantly higher than the concordance between experimental and predicted values. We also show that computational models built from one data set are less predictive for other data sources and highlight the importance of bias correction for assessing selectivity data. Finally, we show that small-molecule target space relationships derived from different data sources and predictive models share overall similarity but can significantly differ in details.

Molecular library design using multi-objective optimization methods.

Advancements in combinatorial chemistry and high-throughput screening technology have enabled the synthesis and screening of large molecular libraries for the purposes of drug discovery. Contrary to initial expectations, the increase in screening library size, typically combined with an emphasis on compound structural diversity, did not result in a comparable increase in the number of promising hits found. In an effort to improve the likelihood of discovering hits with greater optimization potential, more recent approaches attempt to incorporate additional knowledge to the library design process to effectively guide the search. Multi-objective optimization methods capable of taking into account several chemical and biological criteria have been used to design collections of compounds satisfying simultaneously multiple pharmaceutically relevant objectives. In this chapter, we present our efforts to implement a multi-objective optimization method, MEGALib, custom-designed to the library design problem. The method exploits existing knowledge, e.g. from previous biological screening experiments, to identify and profile molecular fragments used subsequently to design compounds compromising the various objectives.

De novo drug design using multiobjective evolutionary graphs.

Drug discovery and development is a complex, lengthy process, and failure of a candidate molecule can occur as a result of a combination of reasons, such as poor pharmacokinetics, lack of efficacy, or toxicity. Successful drug candidates necessarily represent a compromise between the numerous, sometimes competing objectives so that the benefits to patients outweigh potential drawbacks and risks. De novo drug design involves searching an immense space of feasible, druglike molecules to select those with the highest chances of becoming drugs using computational technology. Traditionally, de novo design has focused on designing molecules satisfying a single objective, such as similarity to a known ligand or an interaction score, and ignored the presence of the multiple objectives required for druglike behavior. Recently, methods have appeared in the literature that attempt to design molecules satisfying multiple predefined objectives and thereby produce candidate solutions with a higher chance of serving as viable drug leads. This paper describes the Multiobjective Evolutionary Graph Algorithm (MEGA), a new multiobjective optimization de novo design algorithmic framework that can be used to design structurally diverse molecules satisfying one or more objectives. The algorithm combines evolutionary techniques with graph-theory to directly manipulate graphs and perform an efficient global search for promising solutions. In the Experimental Section we present results from the application of MEGA for designing molecules that selectively bind to a known pharmaceutical target using the ChillScore interaction score family. The primary constraints applied to the design are based on the identified structure of the protein target and a known ligand currently marketed as a drug. A detailed explanation of the key elements of the specific implementation of the algorithm is given, including the methods for obtaining molecular building blocks, evolving the chemical graphs, and scoring the designed molecules. Our findings demonstrate that MEGA can produce structurally diverse candidate molecules representing a wide range of compromises of the supplied constraints and thus can be used as an "idea generator" to support expert chemists assigned with the task of molecular design.

Molecular optimization using computational multi-objective methods.

Improving the profile of a molecule for the drug-discovery process requires the simultaneous optimization of numerous, often competing objectives. Traditionally, standard chemoinformatics methods ignored this problem and focused on the sequential optimization of each single biological or chemical property (ie, a single objective). This approach, known as single-objective optimization (SOOP), strives to discover a single optimal solution to the optimization problem. Implicitly, SOOP-based methods assume that the optimal solution for an objective will also be the optimum for any other objectives involved in the profiling of a molecule. However, when these other objectives are conflicting, as is often the case in drug discovery, the individual optima corresponding to the numerous objectives may vary substantially. Multi-objective optimization (MOOP) methods introduce a new approach for gaining optimality based on compromises and trade-offs among the various objectives. MOOP aims to discover a set of satisfactory compromises that can in turn be used to discover the global optimal solution by optimizing numerous dependent properties simultaneously. MOOP methods have only recently been introduced to the field of chemoinformatics. This paper first presents a brief introduction to issues related to MOOP and then surveys the application of MOOP methods in the field of chemoinformatics.