Research by Guido Capitani

We are interested in protein structure both from a structure determination and from a structural bioinformatics perspective, which means obtaining high resolution "bioimages" of proteins and extracting as much information as possible from them. This information can be useful not only to understand protein structure (structure analysis) but also for structure determination and validation. We are studying protein structures and crystal lattices by employing a variety of approaches, in particular concepts from the field of molecular evolution. We have demonstrated that this approach can be successfully applied to address the problem of distinguishing, in a crystal structure, the biologically relevant interface(s) from those that are merely lattice contacts (reviewed in Capitani et al., 2016). As the first step we used Ka/Ks ratios to detect if a protein-protein interface has a biological character (CRK method, Schärer et al., 2010). In a second phase we used two sequence entropy-based evolutionary indicators and a simple geometric criterion to obtain accurate and efficient classification of protein-protein interfaces (Duarte et al., 2012). This approach (EPPIC, Evolutionary Protein Protein Interface Classifier) is available to the community as an open source command-line tool and as a web server (, linked to a database (EPPIC db) that contains precomputed analysis results (Baskaran et al., 2014) for nearly one million interfaces as of April 2016.

Experimentally, we are involved, in collaboration with the Glockshuber group (ETHZ), in structural studies of a proteinaceous organelle, the type 1 pilus from uropathogenic Escherichia coli (UPEC), and in particular of its main subunit FimA. The type 1 pilus enables UPEC to infect humans by attaching to sugar moieties that reside on the urethral epithelium. It represents a very interesting system to study hybrid approaches to structure determination and structure-function relationships.

Representative publications

Understanding the fabric of protein crystals: computational classification of biological interfaces and crystal contacts.
Capitani G, Duarte JM, Baskaran K, Bliven S and Somody J
Bioinformatics (Oxford) 2016 32:481-9.

"Protein interface classification by evolutionary analysis."
Duarte JM, Srebniak A, Schärer MA and Capitani G
BMC Bioinformatics 2012 13:334.
EPPIC (Evolutionary Protein-Protein Interface Classifier) web server:

"Structural basis of chaperone-subunit complex recognition by the type 1 pilus assembly platform FimD."
Nishiyama M, Horst R, Eidam O, Herrmann T, Ignatov O, Vetsch M, Bettendorff P, Jelesarov I, Grütter MG, Wüthrich K, Glockshuber R and Capitani G
EMBO J 2005 24:2075-86.