Fu, Zheng-QingZheng-QingFuChrzas, JohnJohnChrzasSheldrick, George M.George M.SheldrickRose, JohnJohnRoseWang, B.B.Wang2018-11-072018-11-072007https://resolver.sub.uni-goettingen.de/purl?gro-2/51650A parallel algorithm has been designed for SHELXD to solve the heavy-atom partial structures of protein crystals quickly. Based on this algorithm, a program has been developed to run on high-performance multiple-CPU Linux PCs, workstations or clusters. Tests on the 32-CPU Linux cluster at SER-CAT, APS, Argonne National Laboratory, show that the parallelization dramatically speeds up the process by a factor of roughly the number of CPUs applied, leading to reliable and instant heavy-atom sites solution, which provides the practical opportunity to employ heavy-atom search as an alternative tool for anomalous scattering data quality evaluation during single/multiple-wavelength anomalous diffraction (SAD/MAD) data collection at synchrotron beamlines.review10.1107/S0021889807003998000244926900025