Universiteit Leiden

nl en

Research project

Interacting with proteins, what can we learn?

Supervisor: Gerard van Westen

Gerard van Westen

Suitable for students of: MST master

The availability of public data, more specifically bioactivity data, is steadily increasing. Additionally, the number of crystal structures utilisable for computational studies is accumulating. Traditional methods have leveraged 2D information, like QSAR and Proteochemometrics, due to the increase of noise associated with 3D information. 1 Although Proteochemometrics can describe a binding site based on crystal structures, it typically assumes that the binding site stays relatively constant.2 The goal of this project will be to study models of the Adenosine receptors and incorporate structural information in the form of Interaction Fingerprints into these models. Using this 3D information, derived from crystal structures, will impact the applicability domain of these kinds of models to Virtual Screening.


Our research is interdisciplinary and we have both life scientists and computer scientists working in our group. Accordingly, most projects advertised in the context of the group are suitable for (MSc) students with either a chemical/biological/life science or a computer science/machine learning background. No previous experience in the other field is required, but interest to either get familiar with life science data, or with computational methods, would clearly be an advantage. We strongly support students to publish their results if possible and when the project results are suitable.

  1. Cortes-Ciriano, Isidro, et al. "Applications of proteochemometrics-from species extrapolation to cell line sensitivity modelling."  BMC Bioinformatics 16.Suppl 3 (2015): A4.
  2. van Westen, Gerard JP, et al. "Identifying novel adenosine receptor ligands by simultaneous proteochemometric modeling of rat and human bioactivity data." Journal of Medicinal Chemistry 55.16 (2012): 7010-7020.
  3. Deng, Zhan, Claudio Chuaqui, and Juswinder Singh. "Structural interaction fingerprint (SIFt): a novel method for analyzing three-dimensional protein-ligand binding interactions."  Journal of medicinal chemistry 47.2 (2004): 337-344.
This website uses cookies.  More information.