Chemical fingerprints are very important for both querying chemical data (using similarity in fingerprint space) and the use of molecular entities in machine learning. The RDKit already supports a broad variety of fingerprinting functionality but is limited by the fact that the current fingerprint implementations are all independent and have slightly different APIs. This makes things unnecessarily complicated for users and requires people who are interested in exploring new types of fingerprints to write far more code than it seems like they should have to. In this project we will create a generic decoupled fingerprinting module with a consistent and flexible API.



Boran Adas


  • Andrea Volkamer
  • Nadine Schneider
  • Gregory Landrum