Predicting drug side-effect profiles: a chemical fragment-based approach

Pauwels, Edouard; Stoven, Véronique; Yamanishi, Yoshihiro

doi:10.1186/1471-2105-12-169

Cited by 201 publications

(202 citation statements)

References 28 publications

Supporting

Mentioning

197

Contrasting

Order By: Relevance

“…12 While the utility of chemical features for assessing drug toxicity has been demonstrated earlier, 21,22 here we show their effectiveness on the basis of empirical data of therapeutic side effects by the application of CCA and GCCA models. Knowing the importance of identification of drug features that are critical for specifying their adverse effects, we propose a generalized ordinary canonical correlation analysis model that integrates the target profiles and chemical profiles of drugs.…”

Section: Discussionmentioning

confidence: 98%

Phenotypic side effects prediction by optimizing correlation with chemical and target profiles of drugs

2015

View full text Add to dashboard Cite

Despite technological progresses and improved understanding of biological systems, discovery of novel drugs is an inefficient, arduous and expensive process. Research and development cost of drugs is unreasonably high, largely attributed to the high attrition rate of candidate drugs due to adverse drug reactions. Computational methods for accurate prediction of drug side effects, rooted in empirical data of drugs, have the potential to enhance the efficacy of the drug discovery process. Identification of features critical for specifying side effects would facilitate efficient computational procedures for their prediction. We devised a generalized ordinary canonical correlation model for prediction of drug side effects based on their chemical properties as well as their target profiles. While the former is based on 2D and 3D chemical features, the latter enumerates a systems-level property of drugs. We find that the model incorporating chemical features outperforms that incorporating target profiles. Furthermore we identified the 2D and 3D chemical properties that yield best results, thereby implying their relevance in specifying adverse drug reactions.

show abstract

Section: Discussionmentioning

confidence: 98%

Phenotypic side effects prediction by optimizing correlation with chemical and target profiles of drugs

2015

View full text Add to dashboard Cite

show abstract

“…Similar to the work by Pauwels et al [28], they conducted a large-scale study to develop and validate the ADR prediction model on 1385 known ADRs for 832 FDA (US Food and Drug Administration) approved drugs in SIDER using five machine learning algorithms: LR, Naïve Bayes (NB), KNearest Neighbor (KNN), Random Forest (RF), and SVM. Evaluation results showed that the integration of chemical, biological, and phenotypic properties outperforms the chemical structured-based method (from 0.9054 to 0.9524 with SVM) and has the potential to detect clinically important ADRs at both preclinical and post-market phases for drug surveillance.…”

Section: Integrative Approachmentioning

confidence: 94%

Data mining methodologies for pharmacovigilance

Liu

Matheny

et al. 2012

SIGKDD Explor. Newsl.

View full text Add to dashboard Cite

Medicines are designed to cure, treat, or prevent diseases; however, there are also risks in taking any medicine -particularly short term or long term adverse drug reactions (ADRs) can cause serious harm to patients. Adverse drug events have been estimated to cause over 700,000 emergency department visits each year in the United States. Thus, for medication safety, ADR monitoring is required for each drug throughout its life cycle, including early stages of drug design, different phases of clinical trials, and postmarketing surveillance. Pharmacovigilance (PhV) is the science that concerns with the detection, assessment, understanding and prevention of ADRs. In the pre-marketing stages of a drug, PhV primarily focuses on predicting potential ADRs using preclinical characteristics of the compounds (e.g., drug targets, chemical structure) or screening data (e.g., bioassay data). In the postmarketing stage, PhV has traditionally involved in mining spontaneous reports submitted to national surveillance systems. The research focus is currently shifting toward the use of data generated from platforms outside the conventional framework such as electronic medical records (EMRs), biomedical literature, and patient-reported data in online health forums. The emerging trend of PhV is to link preclinical data from the experimental platform with human safety information observed in the postmarketing phase. This article provides a general overview of the current computational methodologies applied for PhV at different stages of drug development and concludes with future directions and challenges.

show abstract

“…If the j-th element in y new has a high score, the new drug is predicted to have the j-th side-effect (j=1,2,…,q). CCA was showed to be accurate and computationally efficient in prediction of the drug side-effect profiles [20]. Using CCA we augmented the drug-side-effect relationship list with side-effect predictions for drugs that are not included in SIDER, based on their chemical properties.…”

Section: Computing Similarity Of Drug Protein Targetsmentioning

confidence: 99%

Computational Drug Repositioning by Ranking and Integrating Multiple Data Sources

Zhang

Agarwal

Obradović

2013

Advanced Information Systems Engineering

View full text Add to dashboard Cite

Abstract. Drug repositioning helps identify new indications for marketed drugs and clinical candidates. In this study, we proposed an integrative computational framework to predict novel drug indications for both approved drugs and clinical molecules by integrating chemical, biological and phenotypic data sources. We defined different similarity measures for each of these data sources and utilized a weighted k-nearest neighbor algorithm to transfer similarities of nearest neighbors to prediction scores for a given compound. A large margin method was used to combine individual metrics from multiple sources into a global metric. A large-scale study was conducted to repurpose 1007 drugs against 719 diseases. Experimental results showed that the proposed algorithm outperformed similar previously developed computational drug repositioning approaches. Moreover, the new algorithm also ranked drug information sources based on their contributions to the prediction, thus paving the way for prioritizing multiple data sources and building more reliable drug repositioning models.

show abstract

Predicting drug side-effect profiles: a chemical fragment-based approach

Cited by 201 publications

References 28 publications

Phenotypic side effects prediction by optimizing correlation with chemical and target profiles of drugs

Phenotypic side effects prediction by optimizing correlation with chemical and target profiles of drugs

Data mining methodologies for pharmacovigilance

Computational Drug Repositioning by Ranking and Integrating Multiple Data Sources

Contact Info

Product

Resources

About