Modeling of protein structure

Computational workshop on homology modeling of protein structure.

The goal of the workshop is to acquaint the participants with homology modeling of protein structure.

Homology modeling of protein structure is used to obtain a structural model of the protein in absence of experimental data.

Principles of the method: Data for model building is derived from the best homolog with known structure. The quality of the homolog is determined by identity percentage of protein sequences. BLAST online program is used to find the best homologs. Then, an alignment file of modeled protein and template amino acid sequences is built and an executive file for MODELLER program is written. The executive file should contain a reference to alignment file and to coordinate file of the template structure. It is also possible to include optimization methods and assessment of statistical potentials into the executive file. The best models are selected after launching the executive file on base of statistical potential values (for example, discrete optimized protein energy (DOPE) values).

The participants will be acquainted with Protein Data Bank database, structure of coordinate .pdb-files, BLAST search for homologous sequences, homology modeling servers (SWISSModel, Geno3D), protein structure viewers (PyMol, DeepView) and MODELLER program. Every participant will be given a protein sequence for modeling in MODELLER program and on SWISSModel and Geno3D web servers. Options to optimize the models in MODELLER and to assess the derived models by DOPE (discrete optimized protein energy) values will be demonstrated.

Contacts: Abdullatypov Azat Vadimovich. e-mail: