deletions | additions       diff --git a/While_ProCS15_does_not_reproduce__.tex b/While_ProCS15_does_not_reproduce__.tex  index 6d03f2b..0d17611 100644  --- a/While_ProCS15_does_not_reproduce__.tex  +++ b/While_ProCS15_does_not_reproduce__.tex 
...
Comparison of ProCS15 to the empirical methods (CamShift through ShiftX2) generally show considerably lower RMSD for the carbon atoms. The $r$ values are also considerably higher for the empirical methods than for ProCS15 for CA and, especially, CB, while they are comparable for C. The ProCS15 results are quite comparable to the empirical methods for HA and the ProCS15-predicted RMSD values are considerably lower for H compared to CamShift, ppm, and Shaic and but somewhat higher compared to Sparta+ and ShiftX2. Finally, for N the RMSD values for ProCS15 are significantly higher than for the empirical methods while the $r$ values are all comparable.   As mentioned in the introduction the higher RMSD values generally observed for the DFT-based methods compared to the empirical methods is expected as it has been observed many times before (xx-refs). The important issue in the context of structural refinement against measured chemical shifts is  whether the DFT-based methods are more sensitive to relative small differences in structure and therefore more suitable structure. While a thorough investigation of this complex issue  for ProCS15 will be the subject of future studies, next we look at the effect of using different  structural refinement against measured chemical shifts. ensembles on the accuracy next.