This repository contains an analysis of the SIDER 2 (side effect resource) data. This resource automatically extracted concepts from drug labels to create a database of drug side effects and indications.
This analysis provides the source for a companion website, which displays the analysis.
This project is divided into three parts and each part contains a single notebook:
./parse
-- processing the SIDER 2 data in R./similarity
-- calculating compound-compound similarities in R./compounds
-- querying pubchem for compound attributes in python
The SIDER database was described in:
Kuhn M, Campillos M, Letunic I, Jensen LJ, Bork P (2010) A side effect resource to capture phenotypic effects of drugs. Molecular Systems Biology. doi:10.1038/msb.2009.98
To compute compound-compound similarity, we use an analysis similar to:
Campillos M, Kuhn M, Gavin AC, Jensen LJ, Bork P (2008). Drug Target Identification Using Side-Effect Similarity. Science doi:10.1126/science.1158140