Predicting the cytotoxicity of chemicals using ensemble learning methods and molecular fingerprints

Abstract

The prediction of compound cytotoxicity is an important part of the drug discovery process. However, it usually appears as poor predictive performance because the datasets are high‐throughput and have a class‐imbalance problem. In this study, several strategies of performing a structure‐activity relationship study for a cytotoxic endpoint in the AID364 dataset were explored to solve the class‐imbalance problem. Random forest adaboost was used as the base learners for 10 types of molecular fingerprints and an ensemble method and six data‐balancing methods were applied to balance the classes. As a result, the ensemble model using MACCS fingerprint was found to be the best, giving area under the curve of 85.2% ± 0.35%, sensitivity of 81.8% ± 0.65%, and specificity of 76.0% ± 0.12% in fivefold cross‐validation and area under the curve of 78.8%, sensitivity of 55.5% and specificity of 78.5% in external validation. Good performance also appeared on other datasets with different sizes/degrees of imbalance. To explore the structural commonality of cytotoxic compounds, several substructures were identified as an important reference for substructure alerts. The convincing results indicate that the proposed models are helpful in predicting the cytotoxicity of chemicals.

Surface-Related Features Responsible for Cytotoxic Behavior of MXenes Layered Materials Predicted with Machine Learning Approach

“…One such solutions might be the machine learning (ML) procedure. Machine learning has so far proved its applicability for cytotoxicity studies of large number of various chemicals (see Reference [12]) as well as recent predictions of synthesis of various MXenes compounds [13]. In addition, it can be also used to effectively analyse complex surface science data [14].…”

Experimental Comparison of Classification Methods under Class Imbalance

“…The key idea is to train multiple classifiers and then combine them to achieve an overall classification. The ensemble learning approach has been successfully applied to many areas like medical diagnosis [65,66], cheminformatics [67,68], and bioinformatics [69,70]. Ensemble methods reduce the dispersion of model performance and can make reliable prediction performance, but since they are typically based on either sampling and/or cost-sensitive methods as basic classifiers, they may enjoy the benefits of these basic classifiers but also inherit their disadvantages.…”

The Roles of the NLRP3 Inflammasome in Neurodegenerative and Metabolic Diseases and in Relevant Advanced Therapeutic Interventions

“…These days, it is becoming possible to improve cost and time by implementing AI-based methods at different stages of drug design and development. For example, AI is being used in cell classification [154], cell sorting [155], calculating compound properties [156], designing new drug-like molecules [157], computer-aided synthesis of compounds [158,159], predicting the 3D structure of targets, and assay development [160][161][162][163]. These processes are hard to perform but can be optimized and automated by implementing the AI-approaches which could significantly speed up the drug discovery process.…”