Chemotext

Chemotext is a publicly-available Webserver that mines the published literature in PubMed in the form of Medical Subject Headings (MeSH) terms.

The goal of Chemotext is to enable text-based drug-target-disease relationships in order to identify novel drug repurposing candidates and discovery targets.

Text-based relationships have been created between each MeSH term and every article in which the term is referenced.

Through text-mining the published literature, complex connections between MeSH terms in various articles can be made.

Click here to download the tutorial.


There are over 16,000 MeSH Terms, 240,000 Supplementary Records, 19 Million Articles, and 86 Million Relationships in the Chemotext Neo4j Database. Major MeSH updates are released by the National Library of Medicine by the start of every calendar year. The last major Chemotext database update was on November 8, 2017.

Send any feedback to pozefsky@cs.unc.edu. Chemotext is currently public but still in Beta.


Warning: The preferred browser for using Chemotext is Google Chrome. Some features may not be supported on other browsers.

Chemotext was originally conceived by Nancy Baker: Mining connections between chemicals, proteins, and diseases extracted from Medline annotations.

Chemotext is supported by NIH grant 1U01CA207160.