Extracting and codifying study eligibility criteria from ClinicalTrials.gov using cTakes.

This project relies on a private cTAKES module for the time being.

Python 2.7 (probably), with the modules listed in requirements.txt.

We're using MongoDB to cache trial data locally.

If you want to connect to a SMART container, a SMART 0.6+ container is suggested, though there are hacks to support SMART 0.5.

After initializing the submodules, a symlink to the internally used databases should be functional. If not you must create it:

$ cd clinical-trials-gov
$ ln -s ClinicalTrials/databases

There is a SNOMED database at ctakes/ctakes-dictionary-lookup/target/classes/org/apache/ctakes/dictionary/lookup/umls2011ab after checking out cTakes, but it's in HSQLDB format which is not easily usable from Python, so we download and use our own copy:

• download SNOMED CT

• from the directory RF2Release/Full/Terminology place the following files under the given name into the databases directory:

  • sct2_Description_Full-en_INT_xxxxxxx.txt: snomed_desc.csv
  • sct2_Relationship_Full_INT_xxxxxxxx.txt: snomed_rel.csv

  When these files are present, the app will automatically import all SNOMED codes into a local SQLite database, if this has not already been done.

Run the script ClinicalTrials/databases/rxnorm.sh, it will guide you through installing RxNorm.