Numerical Example II Deoxycytidine

The molecular model of deoxycytidine is examined in Tables 2 and 3 to analyze the issues of preconditioning and number of LM-BFGS updates. Energy model details are described elsewhere,22 23 and here it suffices to note that several well-separated local minima correspond to different feasible combinations of the sugar s pseudorotation parameter and the glycosyl (sugar-to-base orientation) dihedral angle. 

The unpreconditioned LM-BFGS runs are not very efficient overall, although their cost per iteration is far better than that of BFGS however, precon- 

In all these deoxycytidine runs, computation time is approximately proportional to the number of function and gradient evaluations. This is typical in molecular computations, as the evaluation and differentiation of the potential energy are expensive and the time-determining factors. The trends observed here extend to biomolecular models and suggest that, with local preconditioning, TN and LM-BFGS are the methods of choice. 

---