labeledin

is a Data Source.

LabeledIn is a human-reviewed, machine-readable, source-linked catalog of labeled indications for human drugs derived from FDA drug labels made available through DailyMed. It links 7,805 drug-disease treatment relationships to UMLS disease concepts and RxNorm drug identifiers, and was extended with a crowdsourced annotation set covering a further 706 drug labels.

Homepage

labeledin

Repository

Unknown

External IDs

No external IDs available

Product Summary

Products

From this Resource
ID Name URL Category Format Description
labeledin.indications LabeledIn Structured Indications LabeledIn_Structured_Results.txt (69.1 KB) Product txt Pipe-delimited structured LabeledIn r...
labeledin.annotated-xml LabeledIn Annotated XML Corpus LabeledIn_annotated_XML.txt (954.1 KB) Product txt LabeledIn annotated corpus in pipe-de...
labeledin.corpus-html LabeledIn Visual HTML Corpus LabeledIn_corpus.html (2.2 MB) Graphical Interface http Single-page HTML view of the LabeledI...
labeledin.crowdsourced LabeledIn Crowdsourced Indications Crowdsourced_Results.txt (351.8 KB) Product txt Crowdsourced LabeledIn extension (201...
labeledin.guidelines LabeledIn Annotation Guidelines Annotation_Guidelines.pdf (267.5 KB) Documentation pdf Annotation guidelines used by the exp...
labeledin.publication LabeledIn Publication j.jbi.2014.08.004 Documentation http Publication describing the LabeledIn ...
From other Resources
ID Name URL Category Format Relation Description
pharmacotherapydb.indications PharmacotherapyDB Indications TSV indications.tsv (75.0 KB) Product tsv was informed by Curated TSV catalog of drug-disease i...

Details

LabeledIn

LabeledIn is a human-reviewed drug indication catalog built by the NCBI text-mining group (Zhiyong Lu lab) from the Indications and Usage sections of FDA drug labels in DailyMed (August 28, 2012 release). Indication mentions were found by automatic disease recognition and then reviewed by expert annotators. Each indication is normalized to a UMLS concept, and each drug label is linked to RxNorm identifiers at the ingredient, ingredient plus dose form, and clinical drug levels. The core set covers 250 drugs frequently accessed in PubMed Health (500 unique DailyMed labels) and contains 7,805 drug-disease treatment relationships, with drugs represented by ingredient, dose form, and strength.

The September 2014 release is distributed from the NCBI FTP site in three formats: a pipe-delimited structured results file, a pipe-delimited file of label text with inline XML annotations, and a visual HTML corpus. Annotation guidelines and overview slides are also provided. A 2015 crowdsourced extension (Crowdsourcing/ subdirectory) covers 706 more drug labels annotated through Amazon Mechanical Turk.

The data carries an NCBI public domain notice (U.S. Government Work). LabeledIn is one of the indication evidence sources aggregated in PharmacotherapyDB.

Is this information incorrect or incomplete? Request an update.

Created: June 02, 2026 | Last modified: November 10, 2026