Proteins initio methods

Semi-empirical methods could thus treat the receptor portion of a single protein molecule as a quantum mechanical region but ab mdio methods cannot. However, both semi-empirical and ab initio methods could treat solvents as a perturbation on a quantum mechanical solute. In the future, HyperChem may have an algorithm for correctly treating the boundary between a classical region and an ab mdio quantum mechanical region in the same molecule. For the time being it does not. 

Although comparative modeling is the most accurate modeling approach, it is limited by its absolute need for a related template structure. For more than half of the proteins and two-thirds of domains, a suitable template structure cannot be detected or is not yet known [9,11]. In those cases where no useful template is available, the ab initio methods are the only alternative. These methods are currently limited to small proteins and at best result only in coarse models with an RMSD error for the atoms that is greater than 4 A. However, one of the most impressive recent improvements in the field of protein structure modeling has occurred in ab initio prediction [155-157]. 

A common way to benefit from the ability to combine different molecular orbital methods in ONIOM is to combine a DFT or ab-initio description of the reactive region with a semi-empirical treatment of the immediate protein environment, including up to 1000 atoms. Due to the requirement for reliable semi-empirical parameters, as discussed in Section 2.2.1, this approach has primarily been used for non-metal or Zn-enzymes. Examples include human stromelysin-1 [83], carboxypeptidase [84], ribonucleotide reductase (substrate reaction) [85], farnesyl transferase [86] and cytosine deaminase [87], Combining two ab-initio methods of different accuracy is not common in biocatalysis applications, and one example from is an ONIOM (MP2 HF) study of catechol O-methyltransferase [88],... 

This paper will discuss the state of the art in 3D structure refinement using empirical, semi-empirical and ab initio methods. We believe that the success story of liquid state NMR in protein structure elucidation is going to continue within the solid state (or membrane environment) if chemical shifts can be successfully exploited. Neutron and X-ray diffraction methods owe their success to a simple formula that connects the measured reflex intensities with the nuclear positions or the electron density, respectively. The better we understand how chemical shifts change with the three-dimensional arrangement of atoms, the more reliably we can construct molecular models from our NMR experiments. As we can in principle determine up to six numbers per nucleus if we perform a full chemical shift tensor analysis, we need to address the question whether whole CS tensor or at least its principal values can be used in structure calculations. 

In 1995 Oldfield reviewed the applications of ab initio methods to chemical shift structure refinement, with special emphasis on the construction of chemical shift hyper-surfaces 3(y>, yf) for proteins, de Dios further discussed the range of ab initio calculations using NMR chemical shifts, focussing on the construction of computed chemical shift hyper-surfaces for proteins. In a more recent review, Ando et alf presented an overview of the interplay... 

Phospholipids located in a membrane, phospholipids (as second messengers) in the cytosol, sphingomyelin, folded protein structures, reactive sites of enzymes, nucleic acids in the zinc finger of the chromosome. Wliat are tire assumptions and constraints for each of these molecular modeling systems (a) Force field methods, (b) Semi-empirical method, (c) Ab initio method. 

Comparison of the accuracy of Fgenesh and Fgenesh+ on the set of human genes that are poorly predicted by ab initio methods, with known protein homologs from other organisms. The set contains 61 genes and 370 exons. CG — percent of correctly predicted genes Sne, Spe are the sensitivity and specificity at the exon level (in %) Snb, Spb are the sensitivity and specificity at the base level (in %), respectively CC — correlation coefficient. 

Fgenesh+ was used to improve the accuracy of prediction for 49 genes. 37 of them were predicted using D. melanogaster s own proteins, already deposited in protein databases. Analysis of these predictions demonstrates that even for such cases prediction of accurate gene structure may not be trivial, although in most cases Fgenesh+ improved the prediction accuracy relative to ab initio methods. 

In accordance with the focus of this chapter on current proteins structure prediction methods, and despite some considerable improvements of ah initio methods (e.g. the fragment assembly approach ROSETTA [37],... 

Ab Initio Methods Predict Localization for All Proteins at Lower Accuracy... 

Several methods have been developed for this purpose. These include (a) comparative or homology modeling, (b) threading method or fold recognition, (c) ab initio method, and (d) visualization of protein folding by computer simulation. 

Ab Initio Method. The ab initio method of prediction of the protein structure is based on thermodynamic considerations that assume that a protein must fold in a manner that the global energy requirement is minimal. This is a difficult approach, but the fact it requires only the amino acid sequence of a protein to model its structure makes it the most feasible approach. This is important particularly in view of the fact that there are more proteins for which only the amino acid sequence is known than of proteins with known structures. 

---