Groups - Basel - Computational Structural Biology - T. Schwede
E-mail Print PDF
Torsten Schwede


The Computational Structural Biology (CSB) group is focusing on the development of methods and algorithms to model, simulate and analyse three-dimensional protein structures and their molecular properties in order to apply these techniques to understand biological processes at a molecular level.


Three-dimensional models of protein structures are key to a detailed understanding of the molecular basis of protein function. Combining sequence with structure information provides invaluable insights for rational strategies to design experiments such as site-directed mutagenesis, studies of disease-related mutations, protein engineering, analysis of ligand specificity, or structure-based drug design. As experimental determination of protein structures by X-ray crystallography and nuclear magnetic resonance (NMR) is time consuming and often technically challenging, experimentally solved 3D structures are only available for a small fraction of the known protein sequences. Therefore, computational approaches are required to close this protein structure knowledge gap. The aim of the CSB group is to develop methods and algorithms to model, simulate and analyse three-dimensional protein structures and their molecular properties, to apply these techniques to understand biological processes
at a molecular level, and to make these methods available to the general biomedical research community.

Projects and Services 

SWISS-MODEL expert system for comparative protein structure modelling

Comparative protein structure modelling (or homology modelling) is exploiting the evolutionary relationship between a target protein of unknown structure and related proteins with known experimental structures. Comparative modelling is currently the most detailed and accurate structure prediction technique available.
With the huge amount of data originating from genome sequencing projects and structural genomics studies, it is necessary to develop expert systems and fully automated methods that are fast and reliable enough to generate accurate models on the scale of whole genomes.
The CSB group develops and operates the SWISS-MODEL expert system for comparative protein structure modelling:
SWISS-MODEL Workspace integrates programs and databases required for interactive homology modelling in a Web-based environment, allowing the user to construct, evaluate and visualize comparative protein models from a computer with Web connection without the need for downloading and installing large program packages and databases.
The SWISS-MODEL Repository is a database of annotated comparative protein structure models. It contains the results of automated large-scale modelling for the sequences from the UniProt database using the SWISS-MODEL pipeline. The integration of structural models with knowledgebases such as UniProt, STRING, or InterPro is central in the computational analysis of sequence-structure-function relationships. We have developed the Protein Model Portal as a one-stop-shop for protein structure information, including both experimental structures and computational models, links to model quality estimation and validation tools, as well as the CAMEO system for Continuous Automated Model Evaluation.


Practical Applications

We are collaborating with research groups in the practical application of protein modelling. Typical examples are rational protein engineering, the design of protein mutations to study protein function or to modify certain protein properties such as substrate specificity.
Molecular modelling provides detailed insights into the interactions governing the specificity of protein-ligand interactions and has had significant impact on structure-based drug design in recent years. Several projects in our group are addressing questions regarding protein-ligand interactions from different perspectives: How can we describe the molecular determinants of receptor specificity? Can we identify small molecules with antiviral properties among commercially available chemical compounds by using virtual screening on a distributed computer GRID to support drug development against neglected tropical diseases? How do errors and inaccuracies of homology models influence the molecular modelling of protein-ligand interactions, and how can these be taken into account in the modelling procedure?

Websites for Further Information

Computational Structural Biology Group:


The Protein Model Portal:


google analytics