Bioinformatic Tools are able to predict a protein's structure either in 2-dimensions (2-D) or three-dimensions (3-D). This may not be a real representation of the actual structure of a protein, however it is a good starting point to predict enzymatic pockets, and protein-protein interaction domains. After prediction, confirmation of protein structure can be made using X-ray Crystallography techniques.
Bioinformatic Programs Used to Predict Secondary Structure:
Experimental data can be used to help in prediction protein structure. The following are experiments that can help in protein structure analysis:
- Localization of disulphide bonds (provide tight restraints on the location of cysteines in space)
- Spectroscopy data which can give clues to the secondary structure composition of your protein of interest
- Site directed mutagenesis experiments, (substituting Alanine A Ala for various amino acids and then checking function of the protein) which can give important insight to the residues which are important in the active site of an enzyme or protein-protein interacting binding sites or domains.
- Obtain experimental data of post-translational modifictions (e.g. phosphorylation and glycosylation sites can suggest residues that must be externally accessible and thus may be surface residues)
- Proteolytic Cleavage sites.
Keeping this data in mind when conducting prediction of protein structure is important, as this is structural data that has been proved experimentally.0
Question whether structure predictions agree with experimental results. This can test the validity of the predicted protein structures.