Mismatch Results

[xy21hb]

Dear all,

I rerun the tutorial for a second time with exactly the same parameters, but the results look so different, any idea?

FXR_12→FXR_85 17.2±0.8 kcal/mol 17.8±0.9 kcal/mol ΔΔG = -0.6±0.9 kcal/mol
FXR_12→FXR_84 -2.0±0.6 kcal/mol -2.7±0.7 kcal/mol ΔΔG = 0.8±0.6 kcal/mol
FXR_12→FXR_88 0.7±378.4 kcal/mol -6.6±1.1 kcal/mol ΔΔG = 7.3±20.5 kcal/mol
FXR_12→FXR_76 -5.8±2.7 kcal/mol -6.1±0.9 kcal/mol ΔΔG = 0.3±1.6 kcal/mol
FXR_12→FXR_74 1.0±0.2 kcal/mol 0.9±0.2 kcal/mol ΔΔG = 0.1±0.2 kcal/mol
================= ΔΔG to FXR_12 ==================
Molecule ΔΔG
FXR_12 0.0 kcal/mol
FXR_74 0.1 kcal/mol
FXR_76 0.3 kcal/mol
FXR_84 0.8 kcal/mol
FXR_85 -0.6 kcal/mol
FXR_88 7.3 kcal/mol

=======================================
FXR_12→FXR_74 1.0±0.0 kcal/mol 1.0±0.0 kcal/mol ΔΔG = 0.0±0.0 kcal/mol
FXR_12→FXR_88 -5.8±0.3 kcal/mol -5.9±0.3 kcal/mol ΔΔG = 0.2±0.3 kcal/mol
FXR_12→FXR_76 -5.7±0.1 kcal/mol -5.3±0.1 kcal/mol ΔΔG = -0.3±0.1 kcal/mol
FXR_12→FXR_84 -3.2±0.1 kcal/mol -2.7±0.1 kcal/mol ΔΔG = -0.6±0.1 kcal/mol
FXR_12→FXR_85 17.3±0.1 kcal/mol 16.7±0.1 kcal/mol ΔΔG = 0.6±0.1 kcal/mol
================= ΔΔG to FXR_12 ==================
Molecule ΔΔG
FXR_12 0.0 kcal/mol
FXR_74 0.0 kcal/mol
FXR_76 -0.3 kcal/mol
FXR_84 -0.6 kcal/mol
FXR_85 0.6 kcal/mol
FXR_88 0.2 kcal/mol

[luancarvalhomartins]

That's expected behavior not related to PyAutoFEP itself, but to the underlying theory of MD sampling. The results will always vary. For some discussion on the statistics of the results, see:

• https://pubs.acs.org/doi/abs/10.1021/acs.jctc.1c00194
• https://pubs.acs.org/doi/10.1021/acs.jcim.0c00900
• https://pubs.acs.org/doi/10.1021/ja512751q