Visualise and Explore Chemical Feature Space of a Dataset

Within this tutorial, we will use the recently developed chemical space mapper, CheS-Mapper ( http://opentox.informatik.uni-freiburg.de/ches-mapper ), to visually explore chemical datasets. We will use the tool for constructing 3D structures, if not already present, and to divide the dataset into clusters of similar compounds and consequently arrange it into 3D space, such that the molecules' spatial proximity reflects their chemical similarity. Using the CheS-Mapper visualization, we will explore the different clusters and highlight the endpoint activity. We will repeat this procedure with extra information computed using various techniques available from chemical libraries. The participants also have the chance to bring their own small dataset in SDF format (up to 100 molecules) and explore it using the CheS-Mapper tool.

Requirements:

latest Java (webstart enabled)

Recommended:

OpenBabel (www.openbabel.org)

Optional:

Rscript (cran.r-project.org)

Dataset including features and endpoint (activity) values:

http://opentox.informatik.uni-freiburg.de/ches-mapper/data/caco2.sdf

http://opentox.informatik.uni-freiburg.de/ches-mapper/data/NCTRER_v4b_232_15Feb2008.ob3d.sdf

Small datasets for structural clustering:

http://opentox.informatik.uni-freiburg.de/ches-mapper/data/chang.sdf

http://opentox.informatik.uni-freiburg.de/ches-mapper/data/krystek.sdf

http://opentox.informatik.uni-freiburg.de/ches-mapper/data/cristalli.sdf

 Complete the evaluation form following this link: Evaluation Form