Prediction of protein structure
Protein structure is understood to mean computer modeling of the spatial (tertiary) structure of a protein, based on its amino acid sequence (primary structure). This task is one of the most important problems in modern bioinformatics. Progress in the development of methods for predicting protein structure is being evaluated within the framework of the CASP World Experiment, which has been conducted every year since 1994 .
types of modeling methods
3 Model structure prediction
4 Protein-free protein prediction
5 See. also
7 External Sources
The Importance of the Problem
In the 1960s, American biochemist, Nobel laureate Christian Anfinsen,  postulated a thermodynamic hypothesis that atoms in protein molecules form naturally. into a thermodynamically stable conformation corresponding to the minimum free energy of the system. In other words, the protein takes a certain spatial form as a result of the constraints dictated by the composition and the physicochemical properties of the amino acids that form it. In turn, protein molecules with a similar spatial structure usually play a similar biological role in cellular level processes. Thus, the structure of a protein can be considered as a transitional masonry between its chemical composition (primary structure) and function. as a Human Genome Project. At the same time, the methods of experimental determination of protein structure are technologically complex, expensive and strongly (by more than two orders of magnitude) lagging behind the methods of determining chemical composition. Thus, as of March 2010, nearly 10 million protein sequences have been deposited in public databases  and this number continues to grow at a rapid rate, while the efforts of major centers of structural genetics worldwide have made the centralized database of PDB protein structures  only fill up to 60,000 structures. It seems that filling the gap between the number of sequences and structures of proteins can only be done by theoretical modeling of the structure of proteins. Nowadays, [When?] There are many examples where computer models of proteins are practically used in biology and medicine, in particular for drug development. Thus, knowledge of protein structure may prompt potential partners for protein interaction, thereby pushing researchers to develop or refine new enzymes or antibodies, or, for example, to explain the phenotypes of mutations carried out or, alternatively, to help identify mutation sites to change specific phenotypes.
Problem Complexity and Types of Modeling Methods
Predicting protein structure is challenging for many reasons. First, the number of possible spatial configurations of proteins is very large. Secondly, the physical principles of protein structure and stability are not yet fully understood. In order to succeed in building a model, structure prediction methods must implement a strategic plan to effectively search the space of possible structures and select the most likely candidates for a native structure. There are currently two major, conceptually different types of methods for narrowing the space of structural conformation searches. proteins: 1) information-based methods based on knowledge derived from experimentally defined structures (knowledge-based methods), and 2) physical methods based on basic Molecular Dynamics Modeling (The Modeling of the First Principles or Ab initio Modeling)
Methods of the first type use the assumption that the protein structure sought may be similar to one or more known protein structures, or at least be composed of elementary structural blocks of such proteins. Methods of the second type do not use information from known structures, but are mainly based on simplified energy potentials and use approximate strategies for finding the minimum of the energy landscape for modeling.
Predicting the structure by model
If among known structures it is possible to find those for which there are grounds claim that they can be somewhat similar to the object of modeling, then they can be used as a template to build a model. This type of method is called template-based modeling. Patterns can be found using direct comparative modeling methods such as BLAST or FASTA, or more complex methods for recognizing structurally similar proteins for weak or virtually non-detectable sequencing (fold recognition or threading methods). The latter group of methods is based on the principle that the structure is evolutionarily more conservative than the sequence, and therefore it is sometimes possible to find related proteins with dissimilar sequences, and then try to "amplify" the sequence of the desired protein through the pattern structure. Theoretically, such proteins can be exposed by constructing and comparing sequential profiles of the desired protein and known structures. protein in this family. With the growth of the structure database, such modeling becomes possible for an increasing number of proteins.
Protein-free protein prediction
When no one of the above methods can find a template, then so-called template-free or de methods are used. new methods). Non-patterned methods include so-called fragmented methods and purely physical methods. The direct simulation of protein conversion by molecular dynamics methods with an energy function that takes into account the details of interactions at the atomic level is nowadays practically insoluble due to the high demand for computational resources. Therefore, most ab initio methods use a simplified atomic structure of proteins.
These methods are less reliable than templates nowadays, but they can sometimes construct models with a general shape close to the native structure of the protein sought. also CASP
↑ Moult, J., et al. (1995). A large-scale experiment to evaluate protein structure prediction methods. Proteins: Structure Function Genetics 23 (3). with. ii – iv.
↑ Anfinsen C (1972). The formation and stabilization of protein structure. Biochem. J. 128 (4). with. 737–49. PMID 4565129.
↑  - External Sources - CASP - Homepage
Наш сайт має систему у функції пошукової системи. Вище: "що ви шукали?"ви можете запитати все в системі з коробкою. Ласкаво просимо в нашу просту, стильну і швидку пошукову систему, яку ми підготували, щоб надати вам найбільш точну та актуальну інформацію.Пошукова система, розроблена для вас, доставляє вам саму актуальну і точну інформацію з простим дизайном і системою швидкого функціонування.
Ви можете знайти майже будь-яку інформацію, яку ви шукаєте на нашому сайті. На даний момент ми служимо тільки англійською, турецькою, російською, українською, казахською та білоруською мовами. Дуже скоро в систему додані нові мови. Життя відомих людей дає вам інформацію, зображення і відео про сотні тем, таких як політики, урядові діячі, лікарі, інтернет-сайти, рослини, технологічні транспортні засоби, автомобілі і т. д.