Target‐Driven Subspace Mapping Methods and Their Applicability Domain Estimation

Soto, Axel J.; Vazquez, Gustavo E.; Strickert, Marc; Ponzoni, Ignacio

doi:10.1002/minf.201100053

Cited by 18 publications

(18 citation statements)

References 43 publications

(43 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, class probability estimates are not broadly applied for setting the AD in chemoinformatics (for exceptions see e.g. [ 18 , 80 ]). Certainly, conformal prediction, which was recently introduced into chemoinformatics [ 24 ], follows a similar philosophy in estimating the reliability of a prediction (by a nonconformity score) for rejecting its prediction if it is too unreliable.…”

Section: Discussionmentioning

confidence: 99%

“…LDA does belong to this class of classifiers. The resulting posterior probability can directly be used as a built-in confidence measure to define the applicability domain [ 18 ]. It is abbreviated as here.…”

Section: Methodsmentioning

confidence: 99%

“…The latter term will be used throughout this contribution to indicate that the scrutinized AD measure actually estimates conditional class probabilities. The latter have first been used explicitly for defining the AD in [ 18 ]. Class probability estimation uses the information of the trained classifier.…”

Section: Methodsmentioning

confidence: 99%

“…These values can be strict probabilities such as the posterior probabilities in linear discriminant analysis or they can be uncalibrated scores. In this case, the only property that holds is that a higher score indicates a higher probability of class membership [ 17 , 18 ]. There are techniques to convert uncalibrated scores into estimated class probabilities [ 19 – 21 ].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Efficiency of different measures for defining the applicability domain of classification models

et al. 2017

View full text Add to dashboard Cite

The goal of defining an applicability domain for a predictive classification model is to identify the region in chemical space where the model’s predictions are reliable. The boundary of the applicability domain is defined with the help of a measure that shall reflect the reliability of an individual prediction. Here, the available measures are differentiated into those that flag unusual objects and which are independent of the original classifier and those that use information of the trained classifier. The former set of techniques is referred to as novelty detection while the latter is designated as confidence estimation. A review of the available confidence estimators shows that most of these measures estimate the probability of class membership of the predicted objects which is inversely related to the error probability. Thus, class probability estimates are natural candidates for defining the applicability domain but were not comprehensively included in previous benchmark studies. The focus of the present study is to find the best measure for defining the applicability domain for a given binary classification technique and to determine the performance of novelty detection versus confidence estimation. Six different binary classification techniques in combination with ten data sets were studied to benchmark the various measures. The area under the receiver operating characteristic curve (AUC ROC) was employed as main benchmark criterion. It is shown that class probability estimates constantly perform best to differentiate between reliable and unreliable predictions. Previously proposed alternatives to class probability estimates do not perform better than the latter and are inferior in most cases. Interestingly, the impact of defining an applicability domain depends on the observed area under the receiver operator characteristic curve. That means that it depends on the level of difficulty of the classification problem (expressed as AUC ROC) and will be largest for intermediately difficult problems (range AUC ROC 0.7–0.9). In the ranking of classifiers, classification random forests performed best on average. Hence, classification random forests in combination with the respective class probability estimate are a good starting point for predictive binary chemoinformatic classifiers with applicability domain.Graphical abstract. Electronic supplementary materialThe online version of this article (doi:10.1186/s13321-017-0230-2) contains supplementary material, which is available to authorized users.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Efficiency of different measures for defining the applicability domain of classification models

et al. 2017

View full text Add to dashboard Cite

show abstract

“…Soto et al [6] focus on estimating a model's applicability domain for virtual compound screening by target-driven subspace mapping. The concept of Correlative Matrix Mapping is introduced to chemoinformatics as a robust projection technique for compound classification with an improved level of confidence.…”

mentioning

confidence: 99%

Editorial: Charting Chemical Space: Challenges and Opportunities for Artificial Intelligence and Machine Learning

Baldi¹,

Müller²,

Schneider³

2011

Molecular Informatics

View full text Add to dashboard Cite

In spite of its central role and position between physics and biology, chemistry has remained in a somewhat backward state of informatics development compared to its two close scientific relatives, primarily for historical reasons. Computers, open public databases, and large collaborative projects have become the pervasive hallmark of research in physics and biology, but are still at a comparably early stage of development in chemistry. Recently, however, data banks containing millions of small molecules have become freely available, and large repositories of chemical reactions have been developed. These data create a wealth of fascinating informatics and machine learning challenges to efficiently store, search, and predict the physical, chemical, and biological properties of small molecules and reactions and chart "chemical space". Profound understanding of its structure and appropriate computational models will have significant scientific and technological impact.

show abstract

Scoring of de novo Designed Chemical Entities by Macromolecular Target Prediction

Button

Hiss

Schneider³

2016

Molecular Informatics

View full text Add to dashboard Cite

Computational de novo molecular design and macromolecular target prediction have become routine in applied cheminformatics. In this study, we have generated populations of drug template-derived designs using ligand-based building block assembly, and predicted their potential targets. The results of our analysis show that the reaction-based de novo design generated new chemical entities with similar properties and pharmacophores as that of the template drugs as well as up to 44 % of the de novo compounds receiving the correct target predictions. Keeping in mind the probabilistic nature of the methods, such a combination of fast and meaningful computational structure generation by reaction-based design and product scoring by target class prediction may be appropriate for prospective application in medicinal chemistry.

show abstract

Target‐Driven Subspace Mapping Methods and Their Applicability Domain Estimation

Cited by 18 publications

References 43 publications

Efficiency of different measures for defining the applicability domain of classification models

Efficiency of different measures for defining the applicability domain of classification models

Editorial: Charting Chemical Space: Challenges and Opportunities for Artificial Intelligence and Machine Learning

Scoring of de novo Designed Chemical Entities by Macromolecular Target Prediction

Contact Info

Product

Resources

About