biotrends

Author: Dario Ghersi

Required module: pymysql

Collection of simple scripts to track the "popularity" of MeSH terms in the biomedical literature.

Setup

MySQL database creation

As MySQL root user, create the medline user and database:

CREATE USER 'medline'@'localhost' IDENTIFIED BY 'pubmed2015';
CREATE DATABASE medline;
GRANT ALL ON medline.* TO 'medline'@'localhost';

Then, as medline user run the createTables.sql script:

mysql --database medline -u medline -p < createTables.sql

Populate the MySQL database

1. Populate the MeSH term and qualifier tables by running the following script on the mesh data:

python fillMeshTable.py mesh/d2015.bin pubmed2015
python fillMeshQualifTable.py mesh/q2015.bin pubmed2015

2. Populate the articles and records tables by running the following script on the (unzipped) Medline XML files:

python fillArticleTable.py medline/medline15n0778.xml

Of course you need to download the full collection of medline XML files, available at:

https://www.nlm.nih.gov/databases/download/pubmed_medline.html

and run the script on each XML file.

Compute the popularity of MeSH terms over the years

The following script computes the number of articles containing given mesh terms:

python extractTimeCourses.py pubmed2015

If you want a denominator to normalize the number of articles published every year, run the following script:

python getNumOfPapersPerYear.py pubmed2015