DocumentCode
2362374
Title
Searching over a domain critical function: Recasting the phase problem of X-ray crystallography
Author
Chang, C.-S. ; DeTitta, G. ; Hauptman, H. ; Miller, R. ; Thuman, P. ; Weeks, C.
Author_Institution
Dept. of Comput. Sci., State Univ. of New York at Buffalo, NY, USA
fYear
1993
fDate
1-5 Mar 1993
Firstpage
183
Lastpage
189
Abstract
This novel technique provides researchers with a powerful and practical tool to solve the phase problem in X-ray crystallography. The authors present a formulation of the phase problem in terms of a minimal function of phases and a new heuristic search algorithm called shake-and-bake for solving this minimal function. This algorithm is based on a hill climbing approach and a back-and-forth transformation between related spaces to utilize implicit chemical information. Since the shake-and-bake algorithm is computationally intensive, several parallel implementations have been explored. These implementations include master/slave multiple instruction/multiple data (MIMD) implementations on an Intel iPSC/860 Hypercube and a Thinking Machines Corporation CM-5. The experimental results indicate that this method is effective on structures, including three previously unknown structures in a variety of space groups
Keywords
X-ray crystallography; chemistry computing; heuristic programming; parallel algorithms; parallel machines; search problems; CM-5; Intel iPSC/860 Hypercube; Thinking Machines Corporation; X-ray crystallography; back-and-forth transformation; chemical information; heuristic search algorithm; hill climbing approach; instruction/multiple data; master/slave multiple; minimal function; parallel implementations; phase problem; shake-and-bake; Antibiotics; Artificial intelligence; Chemicals; Computer science; Crystallography; Heuristic algorithms; Hypercubes; Master-slave; Microorganisms; X-ray diffraction;
fLanguage
English
Publisher
ieee
Conference_Titel
Artificial Intelligence for Applications, 1993. Proceedings., Ninth Conference on
Conference_Location
Orlando, FL
Print_ISBN
0-8186-3840-0
Type
conf
DOI
10.1109/CAIA.1993.366612
Filename
366612
Link To Document