Skip to content

Latest commit

 

History

History
84 lines (59 loc) · 2.79 KB

README.md

File metadata and controls

84 lines (59 loc) · 2.79 KB
Raw

HOW TO BUILD

Clone the mimic-omop repository:

git clone git@github.com:MIT-LCP/mimic-omop.git
cd mimic-omop

All the following commands are assumed to be run in the root path of the repository, i.e. the mimic-omop folder you just cloned and changed into.

Now clone the OMOP Common Data Model DDL to a subfolder. Note that we reset the sub-repository to a specific commit to ensure that the DDL copied is always the same.

git clone https://github.com/OHDSI/CommonDataModel.git
cd CommonDataModel
git reset --hard 0ac0f4bd56c7372dcd3417461a91f17a6b118901
cd ..
cp CommonDataModel/PostgreSQL/*.txt omop/build-omop/postgresql/

Modify the DDL a bit

sed -i 's/^CREATE TABLE \([a-z_]*\)/CREATE UNLOGGED TABLE \1/' "omop/build-omop/postgresql/OMOP CDM postgresql ddl.txt"

Define the PSQL connection parameters you would like to use.

export OMOP_SCHEMA='omop'
export OMOP='host=localhost dbname=postgres user=postgres options=--search_path='$OMOP_SCHEMA
# or, e.g., export OMOP='dbname=mimic options=--search_path='$OMOP_SCHEMA

Build the schema (NOTE: at the moment uses a fixed schema name of omop, edit the script if you need to modify the schema name):

psql "$OMOP" -c "DROP SCHEMA IF EXISTS $OMOP_SCHEMA CASCADE;"
psql "$OMOP" -c "CREATE SCHEMA $OMOP_SCHEMA;"
psql "$OMOP" -f "omop/build-omop/postgresql/OMOP CDM postgresql ddl.txt"

We alter the character columns to text, as there is no performance degradation. This also adds ~4 columns to the NLP table:

psql "$OMOP" -f "omop/build-omop/postgresql/mimic-omop-alter.sql"

We add some comments to the data model:

psql "$OMOP" -f "omop/build-omop/postgresql/omop_cdm_comments.sql"

Symlink your vocabulary folder so that the extras/athena path symlinks to it, e.g.:

ln -s /data/vocab/ extras/athena

Above my /data/vocab/ folder contains CONCEPT_ANCESTOR.csv, CONCEPT_CPT4.csv, CONCEPT.csv, etc.

Note: you can download the vocabulary from here: https://www.ohdsi.org/analytic-tools/athena-standardized-vocabularies/ Don't forget to run the java file to import CPT codes into the concept folder after you download the vocabulary.

Import the vocabulary:

psql "$OMOP" -f "omop/build-omop/postgresql/omop_vocab_load.sql"

(Optional) Indexes may slow down importing of data - so you may want to only build these after running the full ETL.

psql "$OMOP" -f "omop/build-omop/postgresql/OMOP CDM postgresql indexes.txt"

(Optional, experimental) For a similar reason as above, you may want to run this after the full ETL. Note also that since the foreign keys were not used during the construction of the ETL, they may have been violated by the process.

psql "$OMOP" -f "omop/build-omop/postgresql/OMOP CDM postgresql constraints.txt"