Premium
Ab initio structure solution of proteins at atomic resolution using charge‐flipping techniques and cloud computing
Author(s) -
Coelho Alan A.
Publication year - 2021
Publication title -
acta crystallographica section d
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 7.374
H-Index - 138
ISSN - 2059-7983
DOI - 10.1107/s2059798320015090
Subject(s) - charge (physics) , symmetry (geometry) , ab initio , range (aeronautics) , resolution (logic) , atom (system on chip) , computer science , computational physics , cloud computing , physics , computational science , molecular physics , algorithm , atomic physics , materials science , quantum mechanics , artificial intelligence , mathematics , geometry , composite material , embedded system , operating system
Large protein structures at atomic resolution can be solved in minutes using charge‐flipping techniques operating on hundreds of virtual machines (computers) on the Amazon Web Services cloud‐computing platform driven by the computer programs TOPAS or TOPAS‐Academic at a small financial cost. The speed of operation has allowed charge‐flipping techniques to be investigated and modified, leading to two strategies that can solve a large range of difficult protein structures at atomic resolution. Techniques include the use of space‐group symmetry restraints on the electron density as well as increasing the intensity of a randomly chosen high‐intensity electron‐density peak. It is also shown that the use of symmetry restraints increases the chance of finding a solution for low‐resolution data. Finally, a flipping strategy that negates `uranium atom solutions' has been developed for structures that exhibit such solutions during charge flipping.