Dataset Information

Machine learning workflow to enhance predictions of Adverse Drug Reactions (ADRs) through drug-gene interactions: application to drugs for cutaneous diseases.

ABSTRACT: Adverse drug reactions (ADRs) pose critical public health issues, affecting over 6% of hospitalized patients. While knowledge of potential drug-drug interactions (DDI) is necessary to prevent ADR, the rapid pace of drug discovery makes it challenging to maintain a strong insight into DDIs. In this study, we present a novel literature-mining framework for enhancing the predictions of DDIs and ADR types by integrating drug-gene interactions (DGIs). The ADR types were adapted from a DDI corpus, including i) adverse effect; ii) effect at molecular level; iii) effect related to pharmacokinetics; and iv) DDIs without known ADRs. By using random forest classifier our approach achieves an F-score of 0.87 across the ADRs classification using only the DDI features. We then enhanced the performance of the classifier by including DGIs (F-score = 0.90), and applied the classification model trained with the DDI corpus to identify the drugs that might interact with the drugs for cutaneous diseases. We successfully predict previously known ADRs for drugs prescribed to cutaneous diseases, and are also able to identify promising new ADRs.

SUBMITTER: Raja K

PROVIDER: S-EPMC5473874 | biostudies-literature | 2017 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning workflow to enhance predictions of Adverse Drug Reactions (ADRs) through drug-gene interactions: application to drugs for cutaneous diseases.

Raja Kalpana K Patrick Matthew M Elder James T JT Tsoi Lam C LC

Scientific reports 20170616 1

Adverse drug reactions (ADRs) pose critical public health issues, affecting over 6% of hospitalized patients. While knowledge of potential drug-drug interactions (DDI) is necessary to prevent ADR, the rapid pace of drug discovery makes it challenging to maintain a strong insight into DDIs. In this study, we present a novel literature-mining framework for enhancing the predictions of DDIs and ADR types by integrating drug-gene interactions (DGIs). The ADR types were adapted from a DDI corpus, inc ...[more]

PMID: 28623363

Similar Datasets

Project description:ImportanceSerious cutaneous adverse drug reactions (cADRs) are potentially life-threatening drug hypersensitivity reactions involving the skin and internal organs. Antibiotics are a recognized cause of these reactions, but no studies have compared relative risks across antibiotic classes.ObjectivesTo explore the risk of serious cADRs associated with commonly prescribed oral antibiotics, and to characterize outcomes of patients hospitalized for them.Design, setting, and participantsNested case-control study using population-based linked administrative datasets among adults aged 66 years or older who received at least 1 oral antibiotic between 2002 and 2022 in Ontario, Canada. Cases were those who had an emergency department (ED) visit or hospitalization for serious cADRs within 60 days of the prescription, and each case was matched with up to 4 controls who did not.ExposureVarious classes of oral antibiotics.Main outcomes and measuresConditional logistic regression estimate of the association between different classes of oral antibiotics and serious cADRs, using macrolides as the reference group.ResultsDuring the 20-year study period, we identified 21 758 older adults (median age, 75 years; 64.1% female) who had an ED visit or hospitalization for serious cADRs following antibiotic therapy and 87 025 matched controls who did not. In the primary analysis, sulfonamide antibiotics (adjusted odds ratio [aOR], 2.9; 95% CI, 2.7-3.1) and cephalosporins (aOR, 2.6; 95% CI, 2.5-2.8) were most strongly associated with serious cADRs relative to macrolides. Additional associations were evident with nitrofurantoin (aOR, 2.2; 95% CI, 2.1-2.4), penicillins (aOR, 1.4; 95% CI, 1.3-1.5), and fluoroquinolones (aOR, 1.3; 95% CI, 1.2-1.4). The crude rate of ED visits or hospitalization for cADRs was highest for cephalosporins (4.92 per 1000 prescriptions; 95% CI, 4.86-4.99) and sulfonamide antibiotics (3.22 per 1000 prescriptions; 95% CI, 3.15-3.28). Among the 2852 case patients hospitalized for cADRs, the median length of stay was 6 days (IQR, 3-13 days), 9.6% required transfer to a critical care unit, and 5.3% died in the hospital.Conclusion and relevanceCommonly prescribed oral antibiotics are associated with an increased risk of serious cADRs compared with macrolides, with sulfonamides and cephalosporins carrying the highest risk. Prescribers should preferentially use lower-risk antibiotics when clinically appropriate.

Project description:PurposeNonsteroidal anti-inflammatory drugs (NSAIDs) are common cause of severe cutaneous adverse reactions (SCARs). The present study aimed to investigate the characteristics of SCARs induced by NSAIDs in the Korean SCAR registry.MethodsA retrospective survey of NSAID-induced SCARs recorded between 2010 and 2015 at 27 university hospitals in Korea was conducted. Clinical phenotypes of SCARs were classified into Stevens-Johnson syndrome (SJS), toxic epidermal necrolysis (TEN), SJS-TEN overlap syndrome and drug reaction with eosinophilia and systemic symptoms (DRESS). Causative NSAIDs were classified into 7 groups according to their chemical properties: acetaminophen, and propionic, acetic, salicylic, fenamic and enolic acids.ResultsA total of 170 SCARs, consisting of 85 SJS, 32 TEN, 17 SJS-TEN overlap syndrome and 36 DRESS reactions, were induced by NSAIDs: propionic acids (n=68), acetaminophen (n=38), acetic acids (n=23), salicylic acids (n=16), coxibs (n=8), fenamic acids (n=7), enolic acids (n=5) and unclassified (n=5). Acetic acids (22%) and coxibs (14%) accounted for higher portions of DRESS than other SCARs. The phenotypes of SCARs induced by both propionic and salicylic acids were similar (SJS, TEN and DRESS, in order). Acetaminophen was primarily associated with SJS (27%) and was less involved in TEN (10%). DRESS occurred more readily among subjects experiencing coxib-induced SCARs than other NSAID-induced SCARs (62.5% vs. 19.7%, P = 0.013). The mean time to symptom onset was longer in DRESS than in SJS or TEN (19.1 ± 4.1 vs. 6.8 ±1.5 vs. 12.1 ± 3.8 days). SCARs caused by propionic salicylic acids showed longer latency, whereas acetaminophen- and acetic acid-induced SCARs appeared within shorter intervals.ConclusionsThe present study indicates that the phenotypes of SCARs may differ according to the chemical classifications of NSAIDs. To establish the mechanisms and incidences of NSAID-induced SCARs, further prospective studies are needed.

Dataset Information

Machine learning workflow to enhance predictions of Adverse Drug Reactions (ADRs) through drug-gene interactions: application to drugs for cutaneous diseases.

Publications

Machine learning workflow to enhance predictions of Adverse Drug Reactions (ADRs) through drug-gene interactions: application to drugs for cutaneous diseases.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets