In Cheminformatics, a common task is to predict metabolic reactions and pathways. This is usually done by encoding expert knowledge into transformation rules which are generalisations of collections of reactions and work similar to regular expressions. If the left-hand side of a rule is triggered, the transformation described by the rule is translated into a reaction. In this project, we aim to improve the prediction of metabolic pathways by identifying gaps in the transformation rules. We will do this by first identifying reactions that are not covered by transformation rules and then applying methods from machine learning to identify patterns in these reactions. For our experiments, we will use a database of metabolic biodegradation reactions from enviPath, which stores both transformation patterns and reactions. This project is a great opportunity to combine chemistry and machine learning and explore recent research in the area of Cheminformatics.

Duration and Type

  • 12 week summer scholarship in the New Zealand Summer 2021/2022


  • Programming in Java or Python

Supervisor and contact

  • Joerg Wicker
  • Send CV and transcript by mail to Joerg Wicker