Ab initio prediction of the membrane protein structures

Knowledge of the three dimensional structure of the protein is extremely important for a full understanding of its function and molecular proteins interaction. The structure is typically determined experimentally by X-ray crystallography and NMR spectroscopy, unfortunately membrane proteins provide numerous problems for these methods. A possible solution is the computational prediction. Ab initio prediction of three-dimensional models of the membrane proteins is a complex process which cannot use any available protein structure as a general template. There are few softwares that deal with this process and selected four are described in detail in this work. These are two programs for the prediction of transmembrane helical proteins (Rosetta, EVfold_membrane) and two for the prediction of transmembrane beta barrels (EVfold_bb, 3D-SPOT). The main approaches that are used in the prediction of the three-dimensional structure of a protein are inserting short segments of amino acid sequences which are derived from the determined protein structures (Rosetta), using evolutionary information from many other protein sequences (EVfold) and formation of the beta barrel domains based on combining adjacent antiparallel beta chains (3D-SPOT). Every software uses a variety of external programs to address specific..