A fast clustering algorithm for analyzing highly similar compounds of very large libraries

As a result of the recent developments of high-throughput screening in drug discovery, the number of available screening compounds has been growing rapidly. Chemical vendors provide millions of compounds; however, these compounds are highly redundant. Clustering analysis, a technique that groups similar compounds into families, can be used to analyze such redundancy. Many available clustering methods focus on accurate classification of compounds; they are slow and are not suitable for very large compound libraries. Here is described a fast clustering method based on an incremental clustering algorithm and the 2D fingerprints of compounds. This method can cluster a very large data set with millions of compounds in hours on a single computer. A program implemented with this method, called cd-hit-fp, is available from http://chemspace.org.     

Pharmacophore features distributions in different classes of compounds

A pharmacophore analysis approach was used to investigate and compare different classes of compounds relevant to the drug discovery process (specifically, drug molecules, compounds in high throughput screening libraries, combinatorial chemistry building blocks and nondrug molecules). The distributions for a set of pharmacophore features including hydrogen bond acceptors, hydrogen bond donors, negatively ionizable centers, positively ionizable centers and hydrophobic points, were generated and examined. Significant differences were observed between the pharmacophore profiles obtained for the drug molecules and those obtained for the high-throughput screening compounds, which appear to be closely related to the nondrug pharmacophore distribution. It is suggested that the analysis of pharmacophore profiles could be used as an additional tool for the property-based optimization of compound selection and library design processes, thus improving the odds of success in lead discovery projects.     

Process-driven information management system at a biotech company: concept and implementation

While established pharmaceutical companies have chemical information systems in place to manage their compounds and the associated data, new startup companies need to implement these systems from scratch. Decisions made early in the design phase usually have long lasting effects on the expandability, maintenance effort, and costs associated with the information management system. Careful analysis of work and data flows, both inter- and intradepartmental, and identification of existing dependencies between activities are important. This knowledge is required to implement an information management system, which enables the research community to work efficiently by avoiding redundant registration and processing of data and by timely provision of the data whenever needed. This paper first presents the workflows existing at Anadys, then ARISE, the research information management system developed in-house at Anadys. ARISE was designed to support the preclinical drug discovery process and covers compound registration, analytical quality control, inventory management, high-throughput screening, lower throughput screening, and data reporting.     

基于微流控技术的细胞水平高通量药物筛选系统的研究进展

药物筛选是新药研发的关键步骤,创新药物的发现需要采用适当的药物作用靶点对大量化合物样品进行筛选。高通量筛选系统能够实现数千个反应同时测试和分析,大大提高了药物筛选的实验规模和效率。其中基于细胞水平的高通量药物筛选系统因为更加接近人体生理条件,成为主要的筛选模式。而目前发展成熟的高通量细胞筛选系统主要基于多孔板,存在细胞培养条件单一、耗时费力、试剂消耗量大等问题,且较难实现复杂的组合药物筛选。微流控技术作为一种在微米尺度通道中操纵和控制微流体的技术,具有微量、高效、高通量和自动化的优点,能较好地克服多孔板筛选系统的不足,为构建细胞高通量药物筛选系统提供了一种高效、可靠的技术手段。微流控系统在细胞培养材料、芯片结构设计和流体控制方面均可灵活变化,能更好地实现对细胞生长微环境的调控和模拟。文章综述了基于微流控技术的细胞水平高通量药物筛选系统的研究进展,按照不同的微流体操控模式,对基于灌注流、液滴和微阵列的3种类型的微流控细胞筛选系统进行了分类介绍,并分别总结了它们的优缺点,最后展望了微流控细胞水平高通量药物筛选系统的发展前景,提出了该领域目前存在的问题以及解决问题的方向。     

Automatic mass spectrometry method development for drug discovery: application in metabolic stability assays

High-throughput metabolic screening has been requested routinely to keep pace with high-throughput organic synthesis. Liquid chromatography/tandem mass spectrometry (LC/MS/MS) with a fast gradient has become the method of choice for the task due to its sensitivity and selectivity. We have developed an automated system that consists of a robotic system for in vitro incubation and a commercially available software package for automatic MS/MS method development. A short, generic LC gradient and MS conditions that are applicable to most compounds have been developed to minimize the method development time and data analysis. This system has been used to support a number of in vitro screening assays in early drug discovery phase including microsomal stability and protein binding.     

Novel statistical approach for primary high-throughput screening hit selection

The standard activity threshold-based method (the "top X" approach), currently widely used in the high-throughput screening (HTS) data analysis, is ineffective at identifying good-quality hits. We have proposed a novel knowledge-based statistical approach, driven by the hidden structure-activity relationship (SAR) within a screening library, for primary hit selection. Application to an in-house ultrahigh-throughput screening (uHTS) campaign has demonstrated it can directly identify active scaffolds containing valuable SAR information with a greatly improved confirmation rate compared to the standard "top X" method (from 55% to 85%). This approach may help produce high-quality leads and expedite the hit-to-lead process in drug discovery.     

A probabilistic approach to classifying metabolic stability

Metabolic stability is an important property of drug molecules that should-optimally-be taken into account early on in the drug design process. Along with numerous medium- or high-throughput assays being implemented in early drug discovery, a prediction tool for this property could be of high value. However, metabolic stability is inherently difficult to predict, and no commercial tools are available for this purpose. In this work, we present a machine learning approach to predicting metabolic stability that is tailored to compounds from the drug development process at Bayer Schering Pharma. For four different in vitro assays, we develop Bayesian classification models to predict the probability of a compound being metabolically stable. The chosen approach implicitly takes the "domain of applicability" into account. The developed models were validated on recent project data at Bayer Schering Pharma, showing that the predictions are highly accurate and the domain of applicability is estimated correctly. Furthermore, we evaluate the modeling method on a set of publicly available data.     

A generic fast solid-phase extraction high-performance liquid chromatography/mass spectrometry method for high-throughput drug discovery

High-performance liquid chromatography/mass spectrometry (HPLC/MS) is increasingly perceived to be an essential tool in drug discovery at many key steps, like drug screening, lead identification, ADME profiling, and drug metabolism and pharmacology studies. High-throughput screenings in the early phase for metabolic stability, protein binding, permeability (ADME) and bioavailability are widely used to weed out compounds that do not exhibit the necessary characteristics. For such high-throughput LC/MS studies, a generic LC/MS method that can be used for a variety of compounds is desired. In this study, we used a small set of compounds with a wide range of properties to guide method development, and achieved a sample throughput of 1.7 min/sample. Here, we present a generic fast method that achieves good peak separation and peak shape on conventional HPLC systems using a column-switching mechanism for on-line solid-phase extraction (SPE)-HPLC/MS analysis. The method has a linear response range from 1 to 500 nM for the tested compounds. When a larger set of 658 randomly picked small molecules were analyzed using this method, 612 were observed with good signal intensity and HPLC peak shapes. This generic fast SPE-LC/MS method has been used to screen more than 1.5 million compounds repetitively against over 200 protein targets for hit confirmation and semi-quantitation of binding constants from biological assays. Over 7000 different compounds for a variety of protein-binding assays have been studied using this method for quantitative analysis as well.     

Virtual Screening and Structure Generation Applied to Drug Design

The methods of computer-aided drug design can be divided into two categories according to whether or not the structures of receptors are known1, corresponding to two principal strategies: (1) searching the bio-active ligands against virtual combinatorial libraries and calculating the affinity energy between ligand and receptor by docking ; (2) QSAR and 3D-structure data-mining. 3D-QSAR method is now applied widely to drug discovery, but this method is generally limited to refine the structu…     

Analysing the output from primary screening

From a perspective of process knowledge and enhancement, the analysis of the results of biological screening should not be limited to the outcome of specific projects, but additionally encompass a process centric view. Summarising outcomes across multiple projects is a powerful tool to gain a greater understanding of biological screening that will also enable optimisation of the strategy for specific projects or target classes. We have analysed a set of 73,651 compounds with reproducible (confirmed) results from 63 high-throughput screening (HTS) campaigns to reveal the underlying trends in the population of active compounds. We have focused on the overall physico-chemical profile of compound populations derived from biological screening since the in vivo activity of drug molecules is the result of physico-chemical and structural properties of the compound.     

Stalis: A Computational Method for Template-Based Ab Initio Ligand Design

Proteins interact with small molecules through specific molecular recognition, which is central to essential biological functions in living systems. Therefore, understanding such interactions is crucial for basic sciences and drug discovery. Here, we present S tructure t emplate-based a b initio li gand design s olution (Stalis), a knowledge-based approach that uses structure templates from the Protein Data Bank libraries of whole ligands and their fragments and generates a set of molecules (virtual ligands) whose structures represent the pocket shape and chemical features of a given target binding site. Our benchmark performance evaluation shows that ligand structure-based virtual screening using virtual ligands from Stalis outperforms a receptor structure-based virtual screening using AutoDock Vina, demonstrating reliable overall screening performance applicable to computational high-throughput screening. However, virtual ligands from Stalis are worse in recognizing active compounds at the small fraction of a rank-ordered list of screened library compounds than crystal ligands, due to the low resolution of the virtual ligand structures. In conclusion, Stalis can facilitate drug discovery research by designing virtual ligands that can be used for fast ligand structure-based virtual screening. Moreover, Stalis provides actual three-dimensional ligand structures that likely bind to a target protein, enabling to gain structural insight into potential ligands. Stalis can be an efficient computational platform for high-throughput ligand design for fundamental biological study and drug discovery research at the proteomic level. © 2019 Wiley Periodicals, Inc.     

Evolving interpretable structure-activity relationships. 1. Reduced graph queries

A new machine learning method is presented for extracting interpretable structure-activity relationships from screening data. The method is based on an evolutionary algorithm and reduced graphs and aims to evolve a reduced graph query (subgraph) that is present within the active compounds and absent from the inactives. The reduced graph representation enables heterogeneous compounds, such as those found in high-throughput screening data, to be captured in a single representation with the resulting query encoding structure-activity information in a form that is readily interpretable by a chemist. The application of the method is illustrated using data sets extracted from the well-known MDDR data set and GSK in-house screening data. Queries are evolved that are consistent with the known SARs, and they are also shown to be robust when applied to independent sets that were not used in training.     

Advantages of atmospheric pressure photoionization mass spectrometry in support of drug discovery

The performance of the atmospheric pressure photoionization (APPI) technique was evaluated against five sets of standards and drug-like compounds and compared to atmospheric pressure chemical ionization (APCI) and electrospray ionization (ESI). The APPI technique was first used to analyze a set of 86 drug standards with diverse structures and polarities with a 100% detection rate. More detailed studies were then performed for another three sets of both drug standards and proprietary drug candidates. All 60 test compounds in these three sets were detected by APPI with an overall higher ionization efficiency than either APCI or ESI. Most of the non-polar compounds in these three sets were not ionized by APCI or ESI. Analysis of a final set of 201 Wyeth proprietary drug candidates by APPI, APCI and ESI provided an additional comparison of the ionization techniques. The detection rates in positive ion mode were 94% for APPI, 84% for APCI, and 84% for ESI. Combining positive and negative ion mode detection, APPI detected 98% of the compounds, while APCI and ESI detected 91%, respectively. This analysis shows that APPI is a valuable tool for day-to-day usage in a pharmaceutical company setting because it is able to successfully ionize more compounds, with greater structural diversity, than the other two ionization techniques. Consequently, APPI could be considered a more universal ionization method, and therefore has great potential in high-throughput drug discovery especially for open access liquid chromatography/mass spectrometry (LC/MS) applications.     

Monoclonal antibody proteomics: discovery and prevalidation of chronic obstructive pulmonary disease biomarkers in a single step

We define mAb proteomics as the global generation of disease specific antibodies that permit mass screening of biomarkers. An integrated, high-throughput, disease-specific mAb-based biomarker discovery platform has been developed. The approach readily provided new biomarker leads with the focus on large-scale discovery and production of mAb-based, disease-specific clinical assay candidates. The outcome of the biomarker discovery process was a highly specific and sensitive assay, applicable for testing of clinical validation paradigms, like response to treatment or correlation with other clinical parameters. In contrast to MS-based or systems biology-based strategies, our process produced prevalidated clinical assays as the outcome of the discovery process. By re-engineering the biomarker discovery paradigm, the encouraging results presented in this paper clearly demonstrate the efficiency of the mAb proteomics approach, and set the grounds for the next steps of studies, namely, the hunt for candidate biomarkers that respond to drug treatment.     

Integration of NMR and high-throughput screening

NMR-based screening has become a powerful method for the identification and analysis of low-molecular weight organic compounds that bind to protein targets and can be utilized in drug discovery programs. In particular, heteronuclear NMR-based screening can yield information about both the affinity and binding location of potential lead compounds. In addition, heteronuclear NMR-based screening has wide applications in complementing and facilitating conventional high-throughout screening programs. This article will describe several strategies for the integration of NMR-based screening and high-throughput screening. The marriage of these two techniques promises to be of tremendous benefit in the triage of hits that come from HTS, and can aid the medicinal chemist in the identification of quality leads that have high potential for further optimization.