View on GitHub

The DREAM Olfaction Prediction Challenge

A crowd-sourced approach to odor prediction from molecular structure

The Challenge

The DREAM Challenges pose fundamental questions about systems biology and translational medicine. Designed and run by a community of researchers from a variety of organizations, these challenges invite participants to propose solutions — fostering collaboration and building communities in the process. The Olfaction Prediction Challenge was organized by Pablo Meyer to build predictive models based upon data collected and made available by Andreas Keller and Leslie Vosshall. Details and results of the challenge are described here.


All code for training models and eliciting predictions used in the most successful challenge algorithms, as well as associated experimental data, is available in this repository. It also includes Jupyter notebooks for reproducing the content of figures found in the manuscript.

Authors and Contributors

This page is maintained by the @dream-olfaction consortium. Specific questions about the organization or execution of code can be directed to @rgerkin.