Compound Activity Data Research Articles

Benchmarking compound activity prediction for real-world drug discovery applications

Identifying active compounds for target proteins is fundamental in early drug discovery. Recently, data-driven computational methods have demonstrated promising potential in predicting compound activities. However, there lacks a well-designed benchmark to comprehensively evaluate these methods from a practical perspective. To fill this gap, we propose a Compound Activity benchmark for Real-world Applications (CARA). Through carefully distinguishing assay types, designing train-test splitting schemes and selecting evaluation metrics, CARA can consider the biased distribution of current real-world compound activity data and avoid overestimation of model performances. We observed that although current models can make successful predictions for certain proportions of assays, their performances varied across different assays. In addition, evaluation of several few-shot training strategies demonstrated different performances related to task types. Overall, we provide a high-quality dataset for developing and evaluating compound activity prediction models, and the analyses in this work may inspire better applications of data-driven models in drug discovery.

GraphsformerCPI: Graph Transformer for Compound-Protein Interaction Prediction.

Accurately predicting compound-protein interactions (CPI) is a critical task in computer-aided drug design. In recent years, the exponential growth of compound activity and biomedical data has highlighted the need for efficient and interpretable prediction approaches. In this study, we propose GraphsformerCPI, an end-to-end deep learning framework that improves prediction performance and interpretability. GraphsformerCPI treats compounds and proteins as sequences of nodes with spatial structures, and leverages novel structure-enhanced self-attention mechanisms to integrate semantic and graph structural features within molecules for deep molecule representations. To capture the vital association between compound atoms and protein residues, we devise a dual-attention mechanism to effectively extract relational features through .cross-mapping. By extending the powerful learning capabilities of Transformers to spatial structures and extensively utilizing attention mechanisms, our model offers strong interpretability, a significant advantage over most black-box deep learning methods. To evaluate GraphsformerCPI, extensive experiments were conducted on benchmark datasets including human, C. elegans, Davis and KIBA datasets. We explored the impact of model depth and dropout rate on performance and compared our model against state-of-the-art baseline models. Our results demonstrate that GraphsformerCPI outperforms baseline models in classification datasets and achieves competitive performance in regression datasets. Specifically, on the human dataset, GraphsformerCPI achieves an average improvement of 1.6% in AUC, 0.5% in precision, and 5.3% in recall. On the KIBA dataset, the average improvement in Concordance index (CI) and mean squared error (MSE) is 3.3% and 7.2%, respectively. Molecular docking shows that our model provides novel insights into the intrinsic interactions and binding mechanisms. Our research holds practical significance in effectively predicting CPIs and binding affinities, identifying key atoms and residues, enhancing model interpretability.

Systematic identification and characterization of compounds with reliable activity against multiple target proteins from different classes

Polypharmacology provides a basis for the treatment of diseases originating from perturbed target and pathway networks. It depends on the use of compounds with suitable multi-target activity and pharmacological effects resulting from multi-target engagement. In addition to their central role for polypharmacology, multi-target compounds are also of interest for basic research, for example, to better understand their binding characteristics or interaction patterns. While it is not unlikely that active compounds might bind to closely related targets, we have been interested in identifying compounds that are active against distantly or unrelated proteins, representing more unexpected instances of multi-target activity. Therefore, we have carried out a systematic search for compounds with activity against proteins from different classes, termed multi-class compounds (MC-CPDs). To this end, currently available compound activity data from medicinal chemistry were carefully curated. Subsequent analysis of ∼120,000 qualifying compounds with high-confidence target annotations identified 689 MC-CPDs with activity against at least three targets from two different protein classes. The target distribution of these compounds was determined and structural relationships were explored together with activity profiles. MC-CPDs covered many target combinations and established a wealth of compound-based target relationships, some of which could be rationalized by considering ligand binding characteristics of distinct targets. As a part of our study, the collection of newly identified MC-CPDs and associated target information is made publicly available.

Novel quinoxalinone-1,2,3-triazole derivatives as potential antifungal agents for plant anthrax disease: Design, synthesis, antifungal activity and SAR study

A series of quinoxalinone-1,2,3-triazole derivatives have been synthesized and tested for antifungal activity against phytopathogenic fungus through mycelial growth inhibition in vitro, spore germination inhibition in vitro and control effect in vivo. The results showed that the pharmacophore quinoxalinone and 1,2,3-triazole displayed promising antifungal activity. The EC50 value against C. gloeosporioides of 6a-1 was 1.17 μg/mL and the compound 6c-1 showed the good bioactivity against C. fragariae Brooks and P. oryzae Cav., displaying EC50 values of 5.56, 4.62 μg/mL, respectively. The inhibitory of spore germination by 6a-1 was performed with EC50 value of 1.91 μg/mL. In addition, the in vivo bioassay indicated that 6a-1 possessed effective control against Pepper Anthrax Disease in protective treatment. For given activity data of target compounds, structure-activity relationship (SAR) study was operated, hoping for optimizing the target structure to find lead compound with higher activity.

Pathway network-based quantitative modeling of the time-dependent and dose-response anti-inflammatory effect of Reduning Injection.

Traditional Chinese medicine (TCM) has extensive healing effects on inflammatory diseases with few side effects. Reduning injection (RDNI), a TCM prescription composed of Lonicera japonica Thunb., Gardenia jasminoides Ellis. and Artemisia annua L., is wildly used for treating inflammatory diseases. However, the mechanism of action of RDNI, like most TCM prescriptions, is unclear due to the complexity of relationships between components and their curative effects. To develop a universal systems pharmacology protocol for mechanism modeling of TCM and apply it to reveal the real-time anti-inflammatory effect of Reduning Injection. Combined with database mining and references, a regulatory mechanism network of inflammation was constructed. A quantitative model was established afterwards by integrating pharmacokinetic data and two network parameters, namely Network Efficiency and Network Flux. The time-dependent and dose-response relationship of RDNI on the regulation of inflammation was then quantitatively evaluated. ELISA tests were performed to verify the reliability of the model. Three cytokines, namely IL-1β, IL-6 and TNF-α were screened out to be markers for evaluation of the anti-inflammatory effect of RDNI. An HPLC method for the simultaneous determination of 10 RDNI compounds in SD rat plasma was established and then applied to pharmacokinetic studies. Based on compound activity and pharmacokinetic data, the time-dependent effect of RDNI were quantitatively predicted by the pathway network-based modeling procedure. The quantitative model established in this work was successfully applied to predict a TCM prescription's real-time dynamic healing effect after administration. It is qualified to provide novel insights into the time-dependent and dose-effect relationship of drugs in an intricate biological system and new strategies for investigating the detailed molecular mechanisms of TCM.

Identification of potential matrix metalloproteinase-2 inhibitors from natural products through advanced machine learning-based cheminformatics approaches.

Matrix metalloproteinase-2 (MMP-2) is capable of degrading Collage TypeIV in the vascular basement membrane and extracellular matrix. Studies have shown that MMP-2 is tightly associated with the biological behavior of malignant tumors. Therefore, the identification of inhibitors targeting MMP-2 could be effective in treating the disease by maintaining extracellular matrix homeostasis. In the pharmaceutical and biomedical fields, many computational tools are widely used, which improve the efficiency of the whole process to some extent. Apart from the conventional cheminformatics approaches (e.g., pharmacophore model and molecular docking), virtual screening strategies based on machine learning also have promising applications. In this study, we collected 2871 compound activity data against MMP-2 from the ChEMBL database and divided the training and test sets in a 3:1 ratio. Four machine learning algorithms were then selected to construct the classification models, and the best-performing model, i.e., the stacking-based fusion model with the highest AUC value in both training and test datasets, was used for the virtual screening of ZINC database. Next, we screened 17 potential MMP-2 inhibitors from the results predicted by the machine learning model via ADME/T analysis. The interactions between these compounds and the target protein were explored through molecular docking calculations, and the results showed that ZINC712249, ZINC4270723, and ZINC15858504 had lower binding free energies than the co-crystal ligand. To further examine the binding stability of the complexes, we performed molecular dynamics simulations and finally identified these three hits as the most promising natural products for MMP-2 inhibitors.

Past and future of trypanosomatids high-throughput phenotypic screening.

Diseases caused by trypanosomatid parasites affect millions of people mainly living in developing countries. Novel drugs are highly needed since there are no vaccines and available treatment has several limitations, such as resistance, low efficacy, and high toxicity. The drug discovery process is often analogous to finding a needle in the haystack. In the last decades a so-called rational drug design paradigm, heavily dependent on computational approaches, has promised to deliver new drugs in a more cost-effective way. Paradoxically however, the mainstay of these computational methods is data-driven, meaning they need activity data for new compounds to be generated and available in databases. Therefore, high-throughput screening (HTS) of compounds still is a much-needed exercise in drug discovery to fuel other rational approaches. In trypanosomatids, due to the scarcity of validated molecular targets and biological complexity of these parasites, phenotypic screening has become an essential tool for the discovery of new bioactive compounds. In this article we discuss the perspectives of phenotypic HTS for trypanosomatid drug discovery with emphasis on the role of image-based, high-content methods. We also propose an ideal cascade of assays for the identification of new drug candidates for clinical development using leishmaniasis as an example.

X-ray Structure-Based Chemoinformatic Analysis Identifies Promiscuous Ligands Binding to Proteins from Different Classes with Varying Shapes.

(1) Background: Compounds with multitarget activity are of interest in basic research to explore molecular foundations of promiscuous binding and in drug discovery as agents eliciting polypharmacological effects. Our study has aimed to systematically identify compounds that form complexes with proteins from distinct classes and compare their bioactive conformations and molecular properties. (2) Methods: A large-scale computational investigation was carried out that combined the analysis of complex X-ray structures, ligand binding modes, compound activity data, and various molecular properties. (3) Results: A total of 515 ligands with multitarget activity were identified that included 70 organic compounds binding to proteins from different classes. These multiclass ligands (MCLs) were often flexible and surprisingly hydrophilic. Moreover, they displayed a wide spectrum of binding modes. In different target structure environments, binding shapes of MCLs were often similar, but also distinct. (4) Conclusions: Combined structural and activity data analysis identified compounds with activity against proteins with distinct structures and functions. MCLs were found to have greatly varying shape similarity when binding to different protein classes. Hence, there were no apparent canonical binding shapes indicating multitarget activity. Rather, conformational versatility characterized MCL binding.

Activity landscape image analysis using convolutional neural networks

Activity landscapes (ALs) are graphical representations that combine compound similarity and activity data. ALs are constructed for visualizing local and global structure–activity relationships (SARs) contained in compound data sets. Three-dimensional (3D) ALs are reminiscent of geographical maps where differences in landscape topology mirror different SAR characteristics. 3D AL models can be stored as differently formatted images and are thus amenable to image analysis approaches, which have thus far not been considered in the context of graphical SAR analysis. In this proof-of-concept study, 3D ALs were constructed for a variety of compound activity classes and 3D AL image variants of varying topology and information content were generated and classified. To these ends, convolutional neural networks (CNNs) were initially applied to images of original 3D AL models with color-coding reflecting compound potency information that were taken from different viewpoints. Images of 3D AL models were transformed into variants from which one-dimensional features were extracted. Other machine learning approaches including support vector machine (SVM) and random forest (RF) algorithms were applied to derive models on the basis of such features. In addition, SVM and RF models were trained using other features obtained from images through edge filtering. Machine learning was able to accurately distinguish between 3D AL image variants with different topology and information content. Overall, CNNs which directly learned feature representations from 3D AL images achieved highest classification accuracy. Predictive performance for CNN, SVM, and RF models was highest for image variants emphasizing topological elevation. In addition, SVM models trained on rudimentary images from edge filtering classified such images with high accuracy, which further supported the critical role of altitude-dependent topological features for image analysis and predictions. Taken together, the findings of our proof-of-concept investigation indicate that image analysis has considerable potential for graphical SAR exploration to systematically infer different SAR characteristics from topological features of 3D ALs.

Applicability Domains Enhance Application of PPARγ Agonist Classifiers Trained by Drug-like Compounds to Environmental Chemicals.

Peroxisome proliferator activator receptor gamma (PPARγ) agonist activity of chemicals is an indicator of concerned health conditions such as fatty liver and obesity. In silico screening PPARγ agonists based on quantitative structure-activity relationship (QSAR) models could serve as an efficient and pragmatic strategy. Owing to the broad research interests in discovery of PPARγ-targeted drugs, a large amount of PPARγ agonist activity data has been produced in the field of medicinal chemistry, facilitating development of robust QSAR models. In this study, random forest classifiers were developed based on the binary-category data transformed from the heterogeneous PPARγ agonist activity data of drug-like compounds. Coupling with applicability domains, capability of the established classifiers for predicting environmental chemicals was evaluated using two external data sets. Our results demonstrated that applicability domains could enhance application of the developed classifiers to predict environmental PPARγ agonists.

Development of a functional food bar containing coffee

PurposeThe purpose of this paper is to develop a cereal bar supplemented with coffee beverage that has health benefits.Design/methodology/approachSix types of cereal bars containing raisins or prunes with different concentrations of coffee were developed. Acceptability tests and physicochemical characterizations were performed by analysis of moisture, ash, protein, lipids, fibres and carbohydrates. Moreover, the levels of phenolic compounds, the fatty acid profile and thein vitroantioxidant activity were evaluated by the DPPH free radical scavenging and iron-chelating activity methods. The bars were assessed using the check-all-that-apply (CATA) methodology. The phenolic compound and antioxidant activity data were evaluated by analysis of variance. Averages were compared by the Scott–Knott test. To verify the characteristics of the food bar per the attributes cited in CATA, main component analysis was performed using the SensoMaker software.FindingsThe concentration of coffee did not affect the centesimal composition values. The highest percentage of scavenging activity of free radicals was observed in the food bar containing raisins, with a maximum concentration of coffee beverage equivalent to 10 mL. These values were higher in cereal bars containing raisins than in bars containing prunes with the same concentrations of coffee. The acceptance sensory test showed good acceptability ratios, ranging from 74.33 to 85.22 among different food bar formulations.Practical implicationsThe bar consisting raisins and 100 per cent coffee presented high values of protection against oxidative stress, phenolic content and satisfying acceptability, thereby making it a novel possible alternative as a differentiated product with possible health-beneficial effects.Social implicationsThis paper provides a differentiated product, through healthy ingredients, with convenience of purchase, besides having added value and possible health beneficial effects.Originality/valueThe bar consisting of raisins and 100 per cent coffee presented high values of phenolic content and protection against oxidative stress, as well as satisfying acceptability, thereby making it a novel possible alternative as a differentiated product with possible health-beneficial effects.

Application of Bioactivity Profile-Based Fingerprints for Building Machine Learning Models.

The volume of high throughput screening data has considerably increased since the beginning of the automated biochemical and cell-based assays era. This information-rich data source provides tremendous repurposing opportunities for data mining. It was recently shown that biochemical or cell-based assay results can be compiled into so-called high-throughput fingerprints (HTSFPs) as a new type of descriptor describing molecular bioactivity profiles which can be applied in virtual screening, iterative screening, and target deconvolution. However, so far, studies around HTSFPs and machine learning have mainly focused on predicting the outcome of molecules in single high-throughput assays, and no one has reported the modeling of compounds' biochemical assay activities toward a panel of target proteins. In this article, we aim at comparing how our in-house HTSFPs perform at this when combined with multitask deep learning versus the single task support vector machine method both in terms of hit identification and of scaffold hopping potential. Performances obtained from the two HTSFP models were reported with respect to the performances of multitask deep learning and support vector machine models built with the structural descriptors ECFP. Moreover, we investigated the effect of high throughput screening false positives and negatives on the performance of the generated models. Our results showed that the two fingerprints yielded in similar performances and diverse hits with very little overlap, thus demonstrating the orthogonality of bioactivity profile-based descriptors with structural descriptors. Therefore, modeling compound activity data using ECFPs together with HTSFPs increases the scaffold hopping potential of the predictive models.

Data-Driven Exploration of Selectivity and Off-Target Activities of Designated Chemical Probes.

Chemical probes are of central relevance for chemical biology. To unambiguously explore the role of target proteins in triggering or mediating biological functions, small molecules used as probes should ideally be target-specific; at least, they should have sufficiently high selectivity for a primary target. We present a thorough analysis of currently available activity data for designated chemical probes to address several key questions: How well defined are chemical probes? What is their level of selectivity? Is there evidence for additional activities? Are some probes “better” than others? Therefore, highly curated chemical probes were collected and their selectivity was analyzed on the basis of publicly available compound activity data. Different selectivity patterns were observed, which distinguished designated high-quality probes.

Exploring differential evolution for inverse QSAR analysis

Inverse quantitative structure-activity relationship (QSAR) modeling encompasses the generation of compound structures from values of descriptors corresponding to high activity predicted with a given QSAR model. Structure generation proceeds from descriptor coordinates optimized for activity prediction. Herein, we concentrate on the first phase of the inverse QSAR process and introduce a new methodology for coordinate optimization, termed differential evolution (DE), that originated from computer science and engineering. Using simulation and compound activity data, we demonstrate that DE in combination with support vector regression (SVR) yields effective and robust predictions of optimized coordinates satisfying model constraints and requirements. For different compound activity classes, optimized coordinates are obtained that exclusively map to regions of high activity in feature space, represent novel positions for structure generation, and are chemically meaningful.

Synthesis, crystal structure and 3D-QSAR studies of antifungal (bis-)1,2,4-triazole Mannich bases containing furyl and substituted piperazine moieties

A series of novel title compounds 6a-6r and 7a-7c have been synthesized by Mannich reaction of the new triazole Schiff base intermidiates, substituted piperazine and formaldehyde under mild conditions in excellent yields. The crystal structure of compound 6i was determined to show a chair conformation of the piperazine ring and an (E)-configuration of the CN double bond. The bioassay results indicated that most of the newly synthesized compounds exhibited excellent in vitro inhibitory activities and broader spectrum against several plant fungi, and were more effective than the control Triadimefon. Several compounds also displayed favourable in vivo antifungal activities. The relationships between the compound structures and various biological activities were discussed. Furthermore, the CoMFA calculation based on the antifungal activity data of compounds 6 against R. cerealis was carried out to establish a 3D-QSAR model, which revealed that steric and electrostatic fields were two most important factors for contributing the bioactivity of the compounds. The present work will provide significant information for guiding optimization of such new structures to develop novel agrochemicals with higher antifungal activities.

Exploring differential evolution for inverse QSAR analysis

Inverse quantitative structure-activity relationship (QSAR) modeling encompasses the generation of compound structures from values of descriptors corresponding to high activity predicted with a given QSAR model. Structure generation proceeds from descriptor coordinates optimized for activity prediction. Herein, we concentrate on the first phase of the inverse QSAR process and introduce a new methodology for coordinate optimization, termed differential evolution (DE), that originated from computer science and engineering. Using simulation and compound activity data, we demonstrate that DE in combination with support vector regression (SVR) yields effective and robust predictions of optimized coordinates satisfying model constraints and requirements. For different compound activity classes, optimized coordinates are obtained that exclusively map to regions of high activity in feature space, represent novel positions for structure generation, and are chemically meaningful.

Exploring differential evolution for inverse QSAR analysis.

Inverse quantitative structure-activity relationship (QSAR) modeling encompasses the generation of compound structures from values of descriptors corresponding to high activity predicted with a given QSAR model. Structure generation proceeds from descriptor coordinates optimized for activity prediction. Herein, we concentrate on the first phase of the inverse QSAR process and introduce a new methodology for coordinate optimization, termed differential evolution (DE), that originated from computer science and engineering. Using simulation and compound activity data, we demonstrate that DE in combination with support vector regression (SVR) yields effective and robust predictions of optimized coordinates satisfying model constraints and requirements. For different compound activity classes, optimized coordinates are obtained that exclusively map to regions of high activity in feature space, represent novel positions for structure generation, and are chemically meaningful.

ExCAPE-DB: an integrated large scale dataset facilitating Big Data analysis in chemogenomics

Chemogenomics data generally refers to the activity data of chemical compounds on an array of protein targets and represents an important source of information for building in silico target prediction models. The increasing volume of chemogenomics data offers exciting opportunities to build models based on Big Data. Preparing a high quality data set is a vital step in realizing this goal and this work aims to compile such a comprehensive chemogenomics dataset. This dataset comprises over 70 million SAR data points from publicly available databases (PubChem and ChEMBL) including structure, target information and activity annotations. Our aspiration is to create a useful chemogenomics resource reflecting industry-scale data not only for building predictive models of in silico polypharmacology and off-target effects but also for the validation of cheminformatics approaches in general.

Compound Data Mining for Drug Discovery.

In recent years, there has been unprecedented growth in compound activity data in the public domain. These compound data provide an indispensable resource for drug discovery in academic environments as well as in the pharmaceutical industry. To handle large volumes of heterogeneous and complex compound data and extract discovery-relevant knowledge from these data, advanced computational mining approaches are required. Herein, major public compound data repositories are introduced, data confidence criteria reviewed, and selected data mining approaches discussed.

Analyzing compound activity records and promiscuity degrees in light of publication statistics

For the generation of contemporary databases of bioactive compounds, activity information is usually extracted from the scientific literature. However, when activity data are analyzed, source publications are typically no longer taken into consideration. Therefore, compound activity data selected from ChEMBL were traced back to thousands of original publications, activity records including compound, assay, and target information were systematically generated, and their distributions across the literature were determined. In addition, publications were categorized on the basis of activity records. Furthermore, compound promiscuity, defined as the ability of small molecules to specifically interact with multiple target proteins, was analyzed in light of publication statistics, thus adding another layer of information to promiscuity assessment. It was shown that the degree of compound promiscuity was not influenced by increasing numbers of source publications. Rather, most non-promiscuous as well as promiscuous compounds, regardless of their degree of promiscuity, originated from single publications, which emerged as a characteristic feature of the medicinal chemistry literature.