Abstract
When the geometry of large systems is optimized using semiempirical methods, problems are encountered that are not found in smaller systems. In particular, the geometry optimizers conventionally used are all unsuitable for the study of large systems. By modifying the popular BFGS procedure for updating the inverse Hessian matrix, and by modifying the calculation of the step, a method that appears to allow the geometries of large systems to be optimized has been developed. The new method has been applied to the calculation of crambin using PM3, in which the geometry was defined using Cartesian coordinates for the molecular backbone and internal coordinates for the side chains.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
