During Cartesian ΔΔG calculations, I encountered severe energy inconsistencies. #1
Description
core contradictions are:
The initial differences were small: In the first step of initial structure relaxation (pre-relaxation), my results differed from the example file by only about 0.5 REU, which falls within an acceptable normal range of variation.
However, a critical and huge difference emerged: During the subsequent Cartesian relaxation (using -beta_cart), a major discrepancy appeared:
My relaxed structure has an energy of approximately -4900 REU.
The structure provided in the example file has an energy of approximately -5300 REU.
Even more strangely: When I re-relaxed the authors' "example file" structure on my own machine, the resulting energy became -4700 REU (still far from -5300, but closer to my -4900 result).
This issue has caused numerous inconsistencies, including sign reversals, in my subsequent alanine scanning ΔΔG results, preventing me from aligning my outcomes with the article's results.