the SD and SEM of the gmx_MMPBSA

[sf-polaris]

Hi, I run the MD and use 20 frames to calculate the energy decomposition of each residue, in the "decomposition dat" files, I find the SD and SEM are large, but when I calculate the SD using the " decomposition csv" files by myself ， the SD is small. How does the "decomposition dat" file calculate the SD ?

[marioernestovaldes]

please send the files you are using to see if something is wrong

[sf-polaris]

FINAL_DECOMP_MMPBSA.csv
FINAL_DECOMP_MMPBSA2.xlsx

Here are my two output files.
the ".csv" files don't calculate the SD and SEM , I converted the ".dat" files to the ".xlsx" files because GitHub cannot send the ".dat" files, and the ".dat" files calculate the SD and SEM. The SD in the ".dat" files are large. I want to know how does the gmxMMPBSA calculate the SD?

[marioernestovaldes]

The SDs are calculated using propagation of uncertainty rules (https://en.wikipedia.org/wiki/Propagation_of_uncertainty) and SD using arrays of values. We may revisit this at some point in the future though...