Virtual Screening Tasks Research Articles

MMD-DTA: A multi-modal deep learning framework for drug-target binding affinity and binding region prediction.

The prediction of drug-target affinity (DTA) plays a crucial role in drug development and the identification of potential drug targets. In recent years, computer-assisted DTA prediction has emerged as a significant approach in this field. In this study, we propose a multi-modal deep learning framework called MMD-DTA for predicting drug-target binding affinity and binding regions. The model can predict DTA while simultaneously learning the binding regions of drug-target interactions through unsupervised learning. To achieve this, MMD-DTA first uses graph neural networks and target structural feature extraction network to extract multi-modal information from the sequences and structures of drugs and targets. It then utilizes the feature interaction and fusion modules to generate interaction descriptors for predicting DTA and interaction strength for binding region prediction. Our experimental results demonstrate that MMD-DTA outperforms existing models based on key evaluation metrics. Furthermore, external validation results indicate that MMD-DTA enhances the generalization capability of the model by integrating sequence and structural information of drugs and targets. The model trained on the benchmark dataset can effectively generalize to independent virtual screening tasks. The visualization of drug-target binding region prediction showcases the interpretability of MMD-DTA, providing valuable insights into the functional regions of drug molecules that interact with proteins.

Transfer learning across different chemical domains: virtual screening of organic materials with deep learning models pretrained on small molecule and chemical reaction data

Machine learning is becoming a preferred method for the virtual screening of organic materials due to its cost-effectiveness over traditional computationally demanding techniques. However, the scarcity of labeled data for organic materials poses a significant challenge for training advanced machine learning models. This study showcases the potential of utilizing databases of drug-like small molecules and chemical reactions to pretrain the BERT model, enhancing its performance in the virtual screening of organic materials. By fine-tuning the BERT models with data from five virtual screening tasks, the version pretrained with the USPTO–SMILES dataset achieved R2 scores exceeding 0.94 for three tasks and over 0.81 for two others. This performance surpasses that of models pretrained on the small molecule or organic materials databases and outperforms three traditional machine learning models trained directly on virtual screening data. The success of the USPTO–SMILES pretrained BERT model can be attributed to the diverse array of organic building blocks in the USPTO database, offering a broader exploration of the chemical space. The study further suggests that accessing a reaction database with a wider range of reactions than the USPTO could further enhance model performance. Overall, this research validates the feasibility of applying transfer learning across different chemical domains for the efficient virtual screening of organic materials.Scientific contributionThis study verifies the feasibility of applying transfer learning to large language models in different chemical fields to help organic materials perform virtual screening. Through the comparison of transfer learning from different chemical fields to a variety of organic material molecules, the high precision virtual screening of organic materials is realized.

Modern machine‐learning for binding affinity estimation of protein–ligand complexes: Progress, opportunities, and challenges

AbstractStructure‐based drug design is a widely applied approach in the discovery of new lead compounds for known therapeutic targets. In most structure‐based drug design applications, the docking procedure is considered the crucial step. Here, a potential ligand is fitted into the binding site, and a scoring function assesses its binding capability. With the rise of modern machine‐learning in drug discovery, novel scoring functions using machine‐learning techniques achieved significant performance gains in virtual screening and ligand optimization tasks on retrospective data. However, real‐world applications of these methods are still limited. Missing success stories in prospective applications are one reason for this. Additionally, the fast‐evolving nature of the field makes it challenging to assess the advantages of each individual method. This review will highlight recent strides toward improved real world applicability of machine‐learning based scoring, enabling a better understanding of the potential benefits and pitfalls of these functions on a project. Furthermore, a systematic way of classifying machine‐learning based scoring that facilitates comparisons will be presented.This article is categorized under: Data Science > Chemoinformatics Data Science > Artificial Intelligence/Machine Learning Software > Molecular Modeling

BigBind: Learning from Nonstructural Data for Structure-Based Virtual Screening.

Deep learning methods that predict protein-ligand binding have recently been used for structure-based virtual screening. Many such models have been trained using protein-ligand complexes with known crystal structures and activities from the PDBBind data set. However, because PDBbind only includes 20K complexes, models typically fail to generalize to new targets, and model performance is on par with models trained with only ligand information. Conversely, the ChEMBL database contains a wealth of chemical activity information but includes no information about binding poses. We introduce BigBind, a data set that maps ChEMBL activity data to proteins from the CrossDocked data set. BigBind comprises 583 K ligand activities and includes 3D structures of the protein binding pockets. Additionally, we augmented the data by adding an equal number of putative inactives for each target. Using this data, we developed Banana (basic neural network for binding affinity), a neural network-based model to classify active from inactive compounds, defined by a 10 μM cutoff. Our model achieved an AUC of 0.72 on BigBind's test set, while a ligand-only model achieved an AUC of 0.59. Furthermore, Banana achieved competitive performance on the LIT-PCBA benchmark (median EF1% 1.81) while running 16,000 times faster than molecular docking with Gnina. We suggest that Banana, as well as other models trained on this data set, will significantly improve the outcomes of prospective virtual screening tasks.

DSDP: A Blind Docking Strategy Accelerated by GPUs.

Virtual screening, including molecular docking, plays an essential role in drug discovery. Many traditional and machine-learning-based methods are available to fulfill the docking task. However, the traditional docking methods are normally extensively time-consuming, and their performance in blind docking remains to be improved. Although the runtime of docking based on machine learning is significantly decreased, their accuracy is still limited. In this study, we take advantage of both traditional and machine-learning-based methods and present a method, deep site and docking pose (DSDP), to improve the performance of blind docking. For traditional blind docking, the entire protein is covered by a cube, and the initial positions of ligands are randomly generated in this cube. In contrast, DSDP can predict the binding site of proteins and provide an accurate searching shape and initial positions for further conformational sampling. The sampling task of DSDP makes use of the score function and a similar but modified searching strategy of AutoDock Vina, accelerated by implementation in GPUs. We systematically compare its performance in redocking, blind docking, and virtual screening tasks with state-of-the-art methods, including AutoDock Vina, GNINA, QuickVina, SMINA, and DiffDock. In the blind docking task, DSDP reaches a 29.8% top-1 success rate (root-mean-squared deviation < 2 Å) on an unbiased and challenging test dataset with 1.2 s wall-clock computational time per system. Its performances on the DUD-E dataset and the time-split PDBBind dataset used in EquiBind, TANKBind, and DiffDock are also evaluated, presenting a 57.2 and 41.8% top-1 success rate with 0.8 and 1.0 s per system, respectively.

MiDruglikeness: Subdivisional Drug-Likeness Prediction Models Using Active Ensemble Learning Strategies.

The drug development pipeline involves several stages including in vitro assays, in vivo assays, and clinical trials. For candidate selection, it is important to consider that a compound will successfully pass through these stages. Using graph neural networks, we developed three subdivisional models to individually predict the capacity of a compound to enter in vivo testing, clinical trials, and market approval stages. Furthermore, we proposed a strategy combing both active learning and ensemble learning to improve the quality of the models. The models achieved satisfactory performance in the internal test datasets and four self-collected external test datasets. We also employed the models as a general index to make an evaluation on a widely known benchmark dataset DEKOIS 2.0, and surprisingly found a powerful ability on virtual screening tasks. Our model system (termed as miDruglikeness) provides a comprehensive drug-likeness prediction tool for drug discovery and development.

Learning protein-ligand binding affinity with atomic environment vectors

Scoring functions for the prediction of protein-ligand binding affinity have seen renewed interest in recent years when novel machine learning and deep learning methods started to consistently outperform classical scoring functions. Here we explore the use of atomic environment vectors (AEVs) and feed-forward neural networks, the building blocks of several neural network potentials, for the prediction of protein-ligand binding affinity. The AEV-based scoring function, which we term AEScore, is shown to perform as well or better than other state-of-the-art scoring functions on binding affinity prediction, with an RMSE of 1.22 pK units and a Pearson’s correlation coefficient of 0.83 for the CASF-2016 benchmark. However, AEScore does not perform as well in docking and virtual screening tasks, for which it has not been explicitly trained. Therefore, we show that the model can be combined with the classical scoring function AutoDock Vina in the context of Delta-learning, where corrections to the AutoDock Vina scoring function are learned instead of the protein-ligand binding affinity itself. Combined with AutoDock Vina, Delta-AEScore has an RMSE of 1.32 pK units and a Pearson’s correlation coefficient of 0.80 on the CASF-2016 benchmark, while retaining the docking and screening power of the underlying classical scoring function.

Shape-Based Descriptors for Efficient Structure-Based Fragment Growing.

Structure-based fragment growing is one of the key techniques in fragment-based drug design. Fragment growing is commonly practiced based on structural and biophysical data. Computational workflows are employed to predict which fragment elaborations could lead to high-affinity binders. Several such workflows exist but many are designed to be long running noninteractive systems. Shape-based descriptors have been proven to be fast and perform well at virtual-screening tasks. They could, therefore, be applied to the fragment-growing problem to enable an interactive fragment-growing workflow. In this work, we describe and analyze the use of specific shape-based directional descriptors for the task of fragment growing. The performance of these descriptors that we call ray volume matrices (RVMs) is evaluated on two data sets containing protein-ligand complexes. While the first set focuses on self-growing, the second measures practical performance in a cross-growing scenario. The runtime of screenings using RVMs as well as their robustness to three dimensional perturbations is also investigated. Overall, it can be shown that RVMs are useful to prefilter fragment candidates. For up to 84% of the 3299 generated self-growing cases and for up to 66% of the 326 generated cross-growing cases, RVMs could create poses with less than 2 Å root-mean-square deviation to the crystal structure with average query speeds of around 30,000 conformations per second. This opens the door for fast explorative screenings of fragment libraries.

Development of New Methods Needs Proper Evaluation-Benchmarking Sets for Machine Learning Experiments for Class A GPCRs.

New computational approaches for virtual screening applications are constantly being developed. However, before a particular tool is used to search for new active compounds, its effectiveness in the type of task must be examined. In this study, we conducted a detailed analysis of various aspects of preparation of respective data sets for such an evaluation. We propose a protocol for fetching data from the ChEMBL database, examine various compound representations in terms of the possible bias resulting from the way they are generated, and define a new metric for comparing the structural similarity of compounds, which is in line with chemical intuition. The newly developed method is also used for the evaluation of various approaches for division of the data set into training and test set parts, which are also examined in detail in terms of being the source of possible results bias. Finally, machine learning methods are applied in cross-validation studies of data sets constructed within the paper, constituting benchmarks for the assessment of computational methods developed for virtual screening tasks. Additionally, analogous data sets for class A G protein-coupled receptors (100 targets with the highest number of records) were prepared. They are available at http://gmum.net/benchmarks/ , together with script enabling reproduction of all results available at https://github.com/lesniak43/ananas .

Enhancing Acute Oral Toxicity Predictions by using Consensus Modeling and Algebraic Form-Based 0D-to-2D Molecular Encodes

Quantitative structure-activity relationships (QSAR) are introduced to predict acute oral toxicity (AOT), by using the QuBiLS-MAS (acronym for quadratic, bilinear and N-Linear maps based on graph-theoretic electronic-density matrices and atomic weightings) framework for the molecular encoding. Three training sets were employed to build the models: EPA training set (5931 compounds), EPA-full training set (7413 compounds), and Zhu training set (10 152 compounds). Additionally, the EPA test set (1482 compounds) was used for the validation of the QSAR models built on the EPA training set, while the ProTox (425 compounds) and T3DB (284 compounds) external sets were employed for the assessment of all the models. The k-nearest neighbor, multilayer perceptron, random forest, and support vector machine procedures were employed to build several base (individual) models. The base models with REPA-training ≥ 0.75 ( R = correlation coefficient) and MAEEPA-training ≤ 0.5 (MAE = mean absolute error) were retained to build consensus models. As a result, two consensus models based on the minimum operator and denoted as M19 and M22, as well as a consensus model based on the weighted average operator and denoted as M24, were selected as the best ones for each training set considered. According to the applicability domain (AD) analysis performed, model M19 (built on the EPA training set) has MAEtest-AD = 0.4044, MAEProTox-AD = 0.4067 and MAET3DB-AD = 0.2586 on the EPA test set, ProTox external set, and T3DB external set, respectively; whereas model M22 (built on the EPA-full set) and model M24 (built on the Zhu set) present MAEProTox-AD = 0.3992 and MAET3DB-AD = 0.2286, and MAEProTox-AD = 0.3773 and MAET3DB-AD = 0.2471 on the two external sets accounted for, respectively. These outcomes were compared and statistically validated with respect to 14 QSAR methods (e.g., admetSAR, ProTox-II) from the literature. As a result, model M22 presents the best overall performance. In addition, a retrospective study on 261 withdrawn drugs due to their toxic/side effects was performed, to assess the usefulness of prospectively using the QSAR models proposed in the labeling of chemicals. A comparison with regard to the methods from the literature was also made. As a result, model M22 has the best ability of labeling a compound as toxic according to the globally harmonized system of classification and labeling of chemicals. Therefore, it can be concluded that the models proposed, especially model M22, constitute prominent tools for studying AOT, at providing the best results among all the methods examined. A freely available software was also developed to be used in virtual screening tasks ( http://tomocomd.com/apps/ptoxra ).

Chaos-embedded particle swarm optimization approach for protein-ligand docking and virtual screening

BackgroundProtein-ligand docking programs are routinely used in structure-based drug design to find the optimal binding pose of a ligand in the protein’s active site. These programs are also used to identify potential drug candidates by ranking large sets of compounds. As more accurate and efficient docking programs are always desirable, constant efforts focus on developing better docking algorithms or improving the scoring function. Recently, chaotic maps have emerged as a promising approach to improve the search behavior of optimization algorithms in terms of search diversity and convergence speed. However, their effectiveness on docking applications has not been explored. Herein, we integrated five popular chaotic maps—logistic, Singer, sinusoidal, tent, and Zaslavskii maps—into PSOVina^{{mathrm{2LS}}}, a recent variant of the popular AutoDock Vina program with enhanced global and local search capabilities, and evaluated their performances in ligand pose prediction and virtual screening using four docking benchmark datasets and two virtual screening datasets.ResultsPose prediction experiments indicate that chaos-embedded algorithms outperform AutoDock Vina and PSOVina in ligand pose RMSD, success rate, and run time. In virtual screening experiments, Singer map-embedded PSOVina^{{mathrm{2LS}}} achieved a very significant five- to sixfold speedup with comparable screening performances to AutoDock Vina in terms of area under the receiver operating characteristic curve and enrichment factor. Therefore, our results suggest that chaos-embedded PSOVina methods might be a better option than AutoDock Vina for docking and virtual screening tasks. The success of chaotic maps in protein-ligand docking reveals their potential for improving optimization algorithms in other search problems, such as protein structure prediction and folding. The Singer map-embedded PSOVina^{{mathrm{2LS}}} which is named PSOVina-2.0 and all testing datasets are publicly available on https://cbbio.cis.umac.mo/software/psovina.

Convolutional neural network scoring and minimization in the D3R 2017 community challenge.

We assess the ability of our convolutional neural network (CNN)-based scoring functions to perform several common tasks in the domain of drug discovery. These include correctly identifying ligand poses near and far from the true binding mode when given a set of reference receptors and classifying ligands as active or inactive using structural information. We use the CNN to re-score or refine poses generated using a conventional scoring function, Autodock Vina, and compare the performance of each of these methods to using the conventional scoring function alone. Furthermore, we assess several ways of choosing appropriate reference receptors in the context of the D3R 2017 community benchmarking challenge. We find that our CNN scoring function outperforms Vina on most tasks without requiring manual inspection by a knowledgeable operator, but that the pose prediction target chosen for the challenge, Cathepsin S, was particularly challenging for de novo docking. However, the CNN provided best-in-class performance on several virtual screening tasks, underscoring the relevance of deep learning to the field of drug discovery.

Modeling Kinase Inhibition Using Highly Confident Data Sets.

2016 , 56 , 895 - 905 ; www.chembioinf.ro ) aiming to build and measure the performance of ligand-based kinase inhibitor prediction models. Here we analyzed kinase-inhibitor pairs with multiple activity points extracted from the ChEMBL database and identified the main sources of inconsistency. Our results indicate that lower IC50 values are usually less affected by errors and reflect more accurately the structure-activity relationship of the molecules against the target, ideally for quantitative structure-activity relationship studies. Further, we modeled the activities of 104 kinases using unbiased target-specific activity points. The performance of predictors built on extended connectivity fingerprints (ECFP4) and two-dimensional pharmacophore fingerprints (PFPs) are compared by means of tolerance intervals (TIs) (95%/95%) in virtual screening (VS) and classification tasks using external random ( RandSets) and diversity-based ( DivSets) test sets. We found that the two encodings perform superior to each other on different kinases in VS and that PFP models perform consistently better in classifying actives (higher sensitivity). Next, we combined the two encodings into a single one (PFPECFP) and demonstrated that especially in VS (as indicated by the exponential receiver operating curve enrichment metric (eROCE)), for the vast majority of kinases the model performance increased compared with the individual fingerprint models. These findings are highlighted in the more challenging DivSets compared with RandSets. The current paper explores the boundaries of inhibitor predictors for individual kinases to enhance VS and ultimately aid the discovery of novel compounds with desirable polypharmacology.

Virtual screening by a new Clustering-based Weighted Similarity Extreme Learning Machine approach.

Machine learning techniques are becoming popular in virtual screening tasks. One of the powerful machine learning algorithms is Extreme Learning Machine (ELM) which has been applied to many applications and has recently been applied to virtual screening. We propose the Weighted Similarity ELM (WS-ELM) which is based on a single layer feed-forward neural network in a conjunction of 16 different similarity coefficients as activation function in the hidden layer. It is known that the performance of conventional ELM is not robust due to random weight selection in the hidden layer. Thus, we propose a Clustering-based WS-ELM (CWS-ELM) that deterministically assigns weights by utilising clustering algorithms i.e. k-means clustering and support vector clustering. The experiments were conducted on one of the most challenging datasets–Maximum Unbiased Validation Dataset–which contains 17 activity classes carefully selected from PubChem. The proposed algorithms were then compared with other machine learning techniques such as support vector machine, random forest, and similarity searching. The results show that CWS-ELM in conjunction with support vector clustering yields the best performance when utilised together with Sokal/Sneath(1) coefficient. Furthermore, ECFP_6 fingerprint presents the best results in our framework compared to the other types of fingerprints, namely ECFP_4, FCFP_4, and FCFP_6.

DockingApp: a user friendly interface for facilitated docking simulations with AutoDock Vina.

Molecular docking is a powerful technique that helps uncover the structural and energetic bases of the interaction between macromolecules and substrates, endogenous and exogenous ligands, and inhibitors. Moreover, this technique plays a pivotal role in accelerating the screening of large libraries of compounds for drug development purposes. The need to promote community-driven drug development efforts, especially as far as neglected diseases are concerned, calls for user-friendly tools to allow non-expert users to exploit the full potential of molecular docking. Along this path, here is described the implementation of DockingApp, a freely available, extremely user-friendly, platform-independent application for performing docking simulations and virtual screening tasks using AutoDock Vina. DockingApp sports an intuitive graphical user interface which greatly facilitates both the input phase and the analysis of the results, which can be visualized in graphical form using the embedded JMol applet. The application comes with the DrugBank set of more than 1400 ready-to-dock, FDA-approved drugs, to facilitate virtual screening and drug repurposing initiatives. Furthermore, other databases of compounds such as ZINC, available also in AutoDock format, can be readily and easily plugged in.

Extremely Randomized Machine Learning Methods for Compound Activity Prediction.

Speed, a relatively low requirement for computational resources and high effectiveness of the evaluation of the bioactivity of compounds have caused a rapid growth of interest in the application of machine learning methods to virtual screening tasks. However, due to the growth of the amount of data also in cheminformatics and related fields, the aim of research has shifted not only towards the development of algorithms of high predictive power but also towards the simplification of previously existing methods to obtain results more quickly. In the study, we tested two approaches belonging to the group of so-called ‘extremely randomized methods’—Extreme Entropy Machine and Extremely Randomized Trees—for their ability to properly identify compounds that have activity towards particular protein targets. These methods were compared with their ‘non-extreme’ competitors, i.e., Support Vector Machine and Random Forest. The extreme approaches were not only found out to improve the efficiency of the classification of bioactive compounds, but they were also proved to be less computationally complex, requiring fewer steps to perform an optimization procedure.

Iran Virtual Screening (IranVScreen): An Integrated Virtual Screening Interface

Background: Virtual screening (VS), as a computational technique, is being used widely in drug discovery research. One of the widely used VS methods is based on the docking of every ligand structure in a specific macromolecule. Multiple popular VS tools do not provide a graphic user interface (GUI). Objectives: In the present study, we have developed a VS software: IranVScreen, which is an easily operable tool for medicinal chemists and pharmacologists to carry out multiple practical virtual screening tasks. Materials and Methods: The software was developed using Microsoft’s Visual Studio 2012 in visual basic and C++. It integrates Open Babel and Autodock Vina platforms. The software includes an easy to use Persian localized GUI. IranVScreen is composed of three application layers: application layer, class interface layer and software entity layer. Application layer consists of three functional nodes. The software entity layer includes the external tools. Results: The GUI provides the required data to other modules and displays the results to the user. Conclusions: These results describe that IranVScreen provides a very intuitive all-in-one GUI to carry out multiple VS tasks in several mouse clicks with minimal requirement of skill. This software localized in Persian language surpasses the language barrier for novice users.

The influence of the inactives subset generation on the performance of machine learning methods.

BackgroundA growing popularity of machine learning methods application in virtual screening, in both classification and regression tasks, can be observed in the past few years. However, their effectiveness is strongly dependent on many different factors.ResultsIn this study, the influence of the way of forming the set of inactives on the classification process was examined: random and diverse selection from the ZINC database, MDDR database and libraries generated according to the DUD methodology. All learning methods were tested in two modes: using one test set, the same for each method of inactive molecules generation and using test sets with inactives prepared in an analogous way as for training. The experiments were carried out for 5 different protein targets, 3 fingerprints for molecules representation and 7 classification algorithms with varying parameters. It appeared that the process of inactive set formation had a substantial impact on the machine learning methods performance.ConclusionsThe level of chemical space limitation determined the ability of tested classifiers to select potentially active molecules in virtual screening tasks, as for example DUDs (widely applied in docking experiments) did not provide proper selection of active molecules from databases with diverse structures. The study clearly showed that inactive compounds forming training set should be representative to the highest possible extent for libraries that undergo screening.

IVSPlat 1.0: an integrated virtual screening platform with a molecular graphical interface.

BackgroundThe virtual screening (VS) of lead compounds using molecular docking and pharmacophore detection is now an important tool in drug discovery. VS tasks typically require a combination of several software tools and a molecular graphics system. Thus, the integration of all the requisite tools in a single operating environment could reduce the complexity of running VS experiments. However, only a few freely available integrated software platforms have been developed.ResultsA free open-source platform, IVSPlat 1.0, was developed in this study for the management and automation of VS tasks. We integrated several VS-related programs into a molecular graphics system to provide a comprehensive platform for the solution of VS tasks based on molecular docking, pharmacophore detection, and a combination of both methods. This tool can be used to visualize intermediate and final results of the VS execution, while also providing a clustering tool for the analysis of VS results. A case study was conducted to demonstrate the applicability of this platform.ConclusionsIVSPlat 1.0 provides a plug-in-based solution for the management, automation, and visualization of VS tasks. IVSPlat 1.0 is an open framework that allows the integration of extra software to extend its functionality and modified versions can be freely distributed. The open source code and documentation are available at http://kyc.nenu.edu.cn/IVSPlat/.

Estimation of the applicability domain of kernel-based machine learning models for virtual screening.

BackgroundThe virtual screening of large compound databases is an important application of structural-activity relationship models. Due to the high structural diversity of these data sets, it is impossible for machine learning based QSAR models, which rely on a specific training set, to give reliable results for all compounds. Thus, it is important to consider the subset of the chemical space in which the model is applicable. The approaches to this problem that have been published so far mostly use vectorial descriptor representations to define this domain of applicability of the model. Unfortunately, these cannot be extended easily to structured kernel-based machine learning models. For this reason, we propose three approaches to estimate the domain of applicability of a kernel-based QSAR model.ResultsWe evaluated three kernel-based applicability domain estimations using three different structured kernels on three virtual screening tasks. Each experiment consisted of the training of a kernel-based QSAR model using support vector regression and the ranking of a disjoint screening data set according to the predicted activity. For each prediction, the applicability of the model for the respective compound is quantitatively described using a score obtained by an applicability domain formulation. The suitability of the applicability domain estimation is evaluated by comparing the model performance on the subsets of the screening data sets obtained by different thresholds for the applicability scores. This comparison indicates that it is possible to separate the part of the chemspace, in which the model gives reliable predictions, from the part consisting of structures too dissimilar to the training set to apply the model successfully. A closer inspection reveals that the virtual screening performance of the model is considerably improved if half of the molecules, those with the lowest applicability scores, are omitted from the screening.ConclusionThe proposed applicability domain formulations for kernel-based QSAR models can successfully identify compounds for which no reliable predictions can be expected from the model. The resulting reduction of the search space and the elimination of some of the active compounds should not be considered as a drawback, because the results indicate that, in most cases, these omitted ligands would not be found by the model anyway.