Design and validation of an automated method to detect known adverse drug reactions in MEDLINE: a contribution from the EU–ADR project

Avillach, Paul; Dufour, Jean-Charles; Diallo, Gayo; Salvo, Francesco; Joubert, Michel; Thiessard, Frantz; Mougin, Fleur; Trifirò, Gianluca; Fourrier‐Réglat, Annie; Pariente, Alexandre; Fieschi, Marius

doi:10.1136/amiajnl-2012-001083

Cited by 54 publications

(50 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The classifiers in those two studies were constructed based on a set of manually selected ontological and textual features and MeSH terms. Avillach et al 16 used MeSH terms (ie, 'chemically-induced,' 'adverse events') to automate the MEDLINE search to determine if a give drug-SE association was already reported in the literature. There are two perceived limitations in using MeSH terms alone in cancer drug SE extraction: first, cancer drug SE information often appears together with drug-disease treatment information in the same articles.…”

Section: Introductionmentioning

confidence: 99%

Toward creation of a cancer drug toxicity knowledge base: automatically extracting cancer drug—side effect relationships from the literature

Xu¹,

Wang²

2014

J Am Med Inform Assoc

View full text Add to dashboard Cite

Objective A comprehensive and machineunderstandable cancer drug-side effect (drug-SE) relationship knowledge base is important for in silico cancer drug target discovery, drug repurposing, and toxicity predication, and for personalized risk-benefit decisions by cancer patients. While US Food and Drug Administration (FDA) drug labels capture well-known cancer drug SE information, much cancer drug SE knowledge remains buried the published biomedical literature. We present a relationship extraction approach to extract cancer drug-SE pairs from the literature. Data and methods We used 21 354 075 MEDLINE records as the text corpus. We extracted drug-SE cooccurrence pairs using a cancer drug lexicon and a clean SE lexicon that we created. We then developed two filtering approaches to remove drug-disease treatment pairs and subsequently a ranking scheme to further prioritize filtered pairs. Finally, we analyzed relationships among SEs, gene targets, and indications. Results We extracted 56 602 cancer drug-SE pairs. The filtering algorithms improved the precision of extracted pairs from 0.252 at baseline to 0.426, representing a 69% improvement in precision with no decrease in recall. The ranking algorithm further prioritized filtered pairs and achieved a precision of 0.778 for top-ranked pairs. We showed that cancer drugs that share SEs tend to have overlapping gene targets and overlapping indications. Conclusions The relationship extraction approach is effective in extracting many cancer drug-SE pairs from the literature. This unique knowledge base, when combined with existing cancer drug SE knowledge, can facilitate drug target discovery, drug repurposing, and toxicity prediction. BACKGROUND

show abstract

Section: Introductionmentioning

confidence: 99%

Toward creation of a cancer drug toxicity knowledge base: automatically extracting cancer drug—side effect relationships from the literature

Xu¹,

Wang²

2014

J Am Med Inform Assoc

View full text Add to dashboard Cite

show abstract

“…[10] Methods and tools, known as data mining, have been developed over the years to analyse large sets of data. Data mining methods in pharmacovigilance have been used with several goals, [8,9] (i) in order to automate the search of publications concerning ADR in Medline, [11] (ii) to correlate and predict post-marketing adverse drug effects based on screening data from public databases of chemical structures like Pubchem, [12] (iii) to develop new algorithms to detect new or latent multi-drug adverse events in Adverse Event Reporting System database [5] and (iv) to find out new pharmacovigilance signal by mining EHR data. [13,14] For example, in a recently published work, [13] Tatonetti NP et al…”

Section: Introductionmentioning

confidence: 99%

Detection of Drug–Drug Interactions Inducing Acute Kidney Injury by Electronic Health Records Mining

et al. 2015

Self Cite

View full text Add to dashboard Cite

Objective: While risk of Acute Kidney Injury (AKI) is well-documented as adverse effects of some drugs, few studies have assessed relationship between Drug-Drug Interactions (DDI) and AKI. Our objective was to develop an algorithm capable of detecting potential signals on this relationship by mining retrospectively data from an electronic health record. 3 Key Points-The re-use of data from Electronic Health Record (EHR) generated by real clinical activities for pharmacoepidemiological studies is booming, it seemed that DDIs could be identified by mining such databases.-We have developed a new signal detection algorithm using data mining techniques and the RIFLE criteria capable to re-use real care data from EHR in order to identify potential pharmacovigilance signal concerning DDIs and AKI.-This study confirms that real care data from EHRs could be utilized to identify new pharmacovigilance signal concerning DDIs.

show abstract

“…Recently, Avillach et al [13] designed a method to automatically detect known adverse drug reactions in MEDLINE, based on Medical Subject Headings (MeSH) including "descriptors", "supplementary concepts" with subheadings such as "chemically induced", "adverse effects" and "pharmacological action". based discovery for novel therapeutic approaches can be found in [59].…”

Section: Text Mining On Medical Literaturementioning

confidence: 99%

“…To cope with this challenge, on the one hand, methods have been developed to improve document retrieval from biomedical literature databases [78,136,54,13] based discovery for novel therapeutic approaches can be found in [59].…”

Section: Problems and Challengesmentioning

confidence: 99%

“…On the one hand, researches have been developing methods to improve document retrieval from biomedical literature databases [78,136,54,13], such as MEDLINE [3], a bibliographic database of life science literature, maintained by the U.S. National Library of Medicine. Leonard et al [78] proposed a Bayesian approach to mine gene and protein references from MEDLINE abstract and title.…”

Section: Text Mining On Medical Literaturementioning

confidence: 99%

See 1 more Smart Citation

Healthcare data mining from multi-source data

Chen¹

View full text Add to dashboard Cite

The "big data" challenge is changing the way we acquire, store, analyse, and draw conclusions from data. How we effectively and efficiently "mine" the data from possibly multiple sources and extract useful information is a critical question. Increasing research attention has been drawn to healthcare data mining, with an ultimate goal to improve the quality of care. The human body is complex and so too the data collected in treating it. Data noise that is often introduced via the collection process makes building Data Mining models a challenging task.This thesis focuses on the classification tasks of mining healthcare data, with the goal of improving the effectiveness of health risk prediction. In particular, we developed algorithms to address issues identified from real healthcare data, such as feature extraction, heterogeneity, label uncertainty, and large unlabeled data.The three main contributions of this research are as follows. First, we developed a new health index called Personal Health Index (PHI) that scores a person's health status based on the examination records of a given population. Second, we identified the key characteristics of the real datasets and issues that were associated with the data. Third, we developed classification algorithms to cope with those issues, particularly, the label uncertainty and large unlabeled data issues.This research takes one step forward towards scoring personal health based on mining increasingly large health records. Particularly, it pioneers exploring the mining of GHE data and tackles the associated challenges. It is our anticipation that in the near future, more robust data-mining-based health scoring systems will be available for healthcare professionals to understand people's health status and thus improve the quality of care.

show abstract

Design and validation of an automated method to detect known adverse drug reactions in MEDLINE: a contribution from the EU–ADR project

Cited by 54 publications

References 11 publications

Toward creation of a cancer drug toxicity knowledge base: automatically extracting cancer drug—side effect relationships from the literature

Toward creation of a cancer drug toxicity knowledge base: automatically extracting cancer drug—side effect relationships from the literature

Detection of Drug–Drug Interactions Inducing Acute Kidney Injury by Electronic Health Records Mining

Healthcare data mining from multi-source data

Contact Info

Product

Resources

About