 |
2Can Support Portal - Protein Structure
Determining Structures
The most accurate technique used to obtain structure coordinates for proteins is called X-Ray Crystallography. As the name implies, one obtains 3D-coordinates by studying the pattern of diffraction produced when a protein crystal is subjected to bombardment by X-rays. This process however is very costly both in terms of money and time.
The most recent technique developed to determine structure is Nuclear Magnetic Resonance (NMR), this can help determine the structure of protein by studying the properties of the nuclei of atoms under a strong magnetic field. Other techniques involve electron microscopy, in particular cryo-EM, which involves super-freezing of pure protein samples.
Bioinformatics and Structures
One of the challenges that has faced structural biologists for many years has been the prediction of secondary and tertiary structures. In order to predict these accurately one need to know how the physio-chemical properties of each amino acid leads to the formation of bonds between them as well as the angles at which these occur.
Early work carried out by Chou and Fasmann (1974) and Garnier et al. (1978) form the basis for first and second generation predictions of secondary structure. These are by no means to be considered accurate, especially today when there are more than 20,000 known structures in the PDB. However, in their day, they proved invaluable in the understanding of the processes and the complexities involved in making these predictions and attributing assumptions about structures. Their main aim was to produce algorithms that could deduce which parts of a protein sequence would fold as a helix or a sheet. If one could deduce these folds, one could use the information to align similar sequences and infer common secondary structure.
More recent work has lead to the creation of specialised databases such as DSSP (Kabsch and Sander, 1983), and FSSP (Holm and Sander, 1996) (Fold Classification based on structure-structure alignment of Proteins). Work was carried out at the EMBL and EBI, and in effect made use of more refined predictions to align sequences with known secondary structures.
Work carried out in 1990s by Rost and Sander (1993, 1994, 1995) at the EMBL in Heidelberg has lead to much refined knowledge about the topic and the application of this knowledge into services such as DALI which compares s in 2 dimensions and PHP which lead to the creation of the PredictProtein server.
More recently work by James Cuff and Geoff Barton (2000) at the EBI has matured into services such as Jnet and Jpred (1998).
With the advent of cheap and fast computer processors, such as those found in today's PCs and Linux farms, it has become possible to carry out fast and reliable comparisons and alignments in 3D. In effect, aligning the coordinate of known proteins to other related ones, in order to determine more accurate structures and determine the proteins function.
The PDBe Database has a series of tutorials concerning Viewing s, Searching s, Ligand Searches, Lectins and Structural Genomics available here.
A few useful links are, 'Protein of the Month' provided by Interpro and 'Molecule of the Month' provided by the Research Collaboratory for
Structural Bioinformatics (RCSB).
|
|