1990 Fiscal Year Final Research Report Summary
Artifical Intelligence Approach to the Prediction of Three Dimensional Structure of Protein
| Project/Area Number |
63430023
|
|---|
| Research Category |
Grant-in-Aid for General Scientific Research (A)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
生物物性学
|
|---|
| Research Institution | Kyoto University |
|---|
Principal Investigator |
GO Nobuhiro Science, Chemistry, Professor, 理学部, 教授 (50011549)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KATAOKA Yosuke Science, Chemistry, Assistant, 理学部, 助手 (30025407)
|
|---|
| Project Period (FY) |
1988 – 1990
|
| Keywords | Protein / Conformation / Structural elements / Secondary Structure / Dynamics |
|---|
| Research Abstract |
By introducing Kubota's Titan graphic workstation and Sun Micro-system's Sun 3/260 and by constructing therefrom an artificial intelligence system for structural analysis of proteins, we have attained the following results in this research project. Among the problems of predicting 3D structures of proteins, the most difficult is the case where information only of amino acid sequence is available. After 30 years of research is this field it has now been widely believed that the approach based on the physicochemical principle involves essential difficulties. Therefore in this project we have tried an essentially different approach based on the pattern recognition, which in turn is based on the biological evolution of proteins. For this purpose we studied distribution of backbone conformations of fragments of polypeptide chain in the globular proteins. We found a clustered distribution. Based on this property of the distribution a quantitative method of classification of backbone structure is possible.
Among the problems of prediction, the easiest is that of predicting conformational changes associated with a site-directed amino acid substitution. In this case it is demanded that the change is predicted to the accuracy necessary to predict associated change of the function. However, conformational change associated with an amino acid substitution is very small, so small in general to be even smaller than the amplitude of fluctuation in the room temperature. This means that what is to be predicted is the change of dynamic structure. From this point of view, studies are directed to the direction of dynamic structures. As a result, it has been elucidated that protein dynamics has dual aspects, harmonic solid-like aspect and anharmonic liquid-like aspect.
|
