1. Protein structure prediction Anttu Kurttio Ville Pietiläinen

2. Introduction Proteins are one of the most important parts in any biological systems. Understanding the folding of the amino- acid chain to produce functional proteins is essential for studying cellular systems. Fast and accessible methods of solving the 3D structure of a protein are in high demand.

3. Protein structure This topic has been covered several times. Next!

4. Computational methods Ab initio- methods –Laws of physics + amino-acid sequence = protein structure –Computes potential energy functions. –Minimum potential energy is the most stable structure and as such the most likely. –Computationally demanding.

5. Comparative methods Based on the limited amount of possible tertiary structure types. Approximately 2000 different types of protein folds. Comparing the sample to a database of known structures, for example Protein Data Bank.

6. Homology modelling Based on the assumption that homologous (related) proteins fold in a similar fashion. Folding is a highly conserved factor, much more so than amino-acid sequence. Finding a match between two distantly related proteins can be difficult.

7. Protein threading Based on the assumption that similar folding has already been found. Comparing parts of the sequence to a database of known three dimensional structures using a scoring function. Works at least somewhat on approximately 80% of new protein sequences.

8. Programs A lot of free programs are available. Server based programs do the computational work. For example Swiss- Model, Rosetta or PSIPRED. Downloadable applications are used for viewing the results. For example Swiss- PdbViewer or Rasmol. Distributed computing promises increases in computational capacity.

9. DEMO Swiss-Model in four easy steps. swissmodel.expasy.org

10. Step 1: Send in the sequence

11. Step 2: Coffee break

12. Step 3: Recieve mail

13. Step 4: Open PDB-file

14. Rejoice! Study your brand new model of a protein.