Dear Dr Case,
Thank you for your answer, as always...
> The simple force field codes used in Amber cannot handle situations where
> the angle is 180: there is no way to compute the force for such an angle.
Yes I understand that, but here this is about 'single point energy
measurement': there is no energy derivative and force defined...
PyRED computes 'single point energy measurement' versus sander to
validate the data; Yes I forgot to underline that in my previous
message...
(by now REDServer Dev. uses AmberTools/sander to run geometry
optimization and compute dE(MM) versus dE(QM) for conformations, but
this is not related to this discussion).
> The force constants for such angles should be zero, at least in my
> understanding.
The FF constants for such angles are not zero in the frcmod file;
values between 40/45 (gaff2) - 50 (parm10) kcal/mol/rad-2 are used. In
the job presented in the former email, P2875, the FF used was 'parm10'
which is internally stored in PyRED.
> You say "sander|pmemd do not find '0'". Is this because
> some angles do not have zero force constants?
The FF constants are not zero as I just wrote; but if Theta(eq) = 180
and measure_theta = 180 the corresponding angle difference = 0, so the
energy = 0.
> At the very end of your
> message, there seem to be some non-zero force constants. Are those in
> parm10, or from somewhere else?
Yes the FF constants come from parm10.
> Do you know what GAFF does it this case?
Ok - I ran a new job, P3095, with GAFF - see below; similar problems
for 'sp' atoms, only...
Best Francois
The corresponding data are only conserved 5 days on the R.E.D. Server
Development.
Below data with the latest gaff2 version from amber26:
https://q4mdfft:q4mdfft@cluster.q4md-forcefieldtools.org/~ucpublic1/ADF2yADF10ADFagXGby3HYyWTTCkvA/P3095.php
Display the Project.config-ArchiveFile and Project.config files... 
#_______Keyword(s) added in the Project.config file by the user
before the input submission procedure:
MOLECULE1-TITLE = CH4_methane
MOLECULE2-TITLE = H2C≈O_methanal
MOLECULE3-TITLE = HC≡N_cyanhydric-acid
MOLECULE4-TITLE = HC≡CH_ethyne
MOLECULE1-ATMTYPE = c3 hc hc hc hc
MOLECULE2-ATMTYPE = c ha ha o
MOLECULE3-ATMTYPE = c1 ha n1
MOLECULE4-ATMTYPE = c1 ha c1 ha
No problem with molecule 1 (methane; 'sp3' angles) or molecule 2
(methanal; 'sp2' angles): single point energy measures are ok.
Problems again with triple bonds: molecule 3 (cyanhydric-acid; 1 'sp'
angle) and molecule 4 (ethyne; 2 'sp' angles)
The simplified/checked 'frcmod.known' file generated by PyRED from gaff2:
https://q4mdfft:q4mdfft.cluster.q4md-forcefieldtools.org/~ucpublic1/ADF2yADF10ADFagXGby3HYyWTTCkvA/P3095/Data-R.E.D.Server/Data-Default-Proj/frcmod.known
c1-c1-ha 40.0 179.2800 adapted from frcmod.user
41.05 Theta(eq) NOT 180.000 ???
ha-c1-n1 45.0 180.0000 adapted from frcmod.user
44.6 Theta(eq) = 180.000 THIS TIME ???
Differences in energies 'sander vs PyRED':
https://q4mdfft:q4mdfft.cluster.q4md-forcefieldtools.org/~ucpublic1/ADF2yADF10ADFagXGby3HYyWTTCkvA/P3095/dE_sander-vs-PyRED_SinglePointEnergyValues.log
3 Mol_mm3-c1.log BND= 0.6695 ANG= 0.0900 etc... sander 2024
4 Mol_mm4-c1.log BND= 0.4776 ANG= 0.0827 etc... sander 2024
3 Energy_mm3-c1.txt BND= 0.6695 ANG= 0.0000 etc... PyRED
finds ANG=0.0000 because Theta(eq) = 180.0000; this is correct
see
https://q4mdfft:q4mdfft.cluster.q4md-forcefieldtools.org/~ucpublic1/ADF2yADF10ADFagXGby3HYyWTTCkvA/P3095/Data-R.E.D.Server/Mol_MM/Energy_mm3-c1.txt
4 Energy_mm4-c1.txt BND= 0.4776 ANG= 0.0126 etc... PyRED
finds ANG=0.0126 because Theta(eq) = 179.2800_???; this is also correct
see
https://q4mdfft:q4mdfft.cluster.q4md-forcefieldtools.org/~ucpublic1/ADF2yADF10ADFagXGby3HYyWTTCkvA/P3095/Data-R.E.D.Server/Mol_MM/Energy_mm4-c1.txt
and manual check outside PyRED:
[xxxx.master P3095] bash z-Measure--AngleEnergy.sh
Usage: z-Measure--AngleEnergy.sh <angle> <Theta0> <force_constant>
[xxxx.master P3095] bash z-Measure--AngleEnergy.sh 180 179.28 40
0.0063 kcal/mol
there are 2 angles in ethyne: 0.0063 + 0.0063 = 0.0126
sander vs PyRED:
--------------------------------------
Mol_mm3-c1.log vs Energy_mm3-c1.txt
kcal/mol
BOND sander vs PyRED single-point energy abs(differences) =
0 [OK]
ANGLE sander vs PyRED single-point energy abs(differences) =
.0900 [WARNING] H-C-N
DIH+IMP sander vs PyRED single-point energy abs(differences) =
0 [OK] (*)
VDW sander vs PyRED single-point energy abs(differences) =
0 [OK]
VDW14 sander vs PyRED single-point energy abs(differences) =
0 [OK]
EEL sander vs PyRED single-point energy abs(differences) =
0 [OK]
EEL14 sander vs PyRED single-point energy abs(differences) =
0 [OK]
--------------------------------------
Mol_mm4-c1.log vs Energy_mm4-c1.txt
kcal/mol
BOND sander vs PyRED single-point energy abs(differences) =
0 [OK]
ANGLE sander vs PyRED single-point energy abs(differences) =
.0701 [WARNING] 2 H-C-C
DIH+IMP sander vs PyRED single-point energy abs(differences) =
0 [OK] (*)
VDW sander vs PyRED single-point energy abs(differences) =
0 [OK]
VDW14 sander vs PyRED single-point energy abs(differences) =
0 [OK]
EEL sander vs PyRED single-point energy abs(differences) =
0 [OK]
EEL14 sander vs PyRED single-point energy abs(differences) =
0 [OK]
Sander 2024 results:
molecule 3 - cyanhydric-acid
https://cluster.q4md-forcefieldtools.org/~ucpublic1/ADF2yADF10ADFagXGby3HYyWTTCkvA/P3095/Data-R.E.D.Server/Mol_MM/MM_sander_vs_pyred_SinglePoint/Mol_mm3-c1.log
i.e. for NSTEP = 1
BOND = 0.6696 ANGLE = 0.0900 DIHED = 0.0000
VDWAALS = 0.0000 EEL = 0.0000 HBOND = 0.0000
1-4 VDW = 0.0000 1-4 EEL = 0.0000 RESTRAINT = 0.0000
ANGLE energy should equal 0.0000 !!!
molecule 4: - ethyne
https://cluster.q4md-forcefieldtools.org/~ucpublic1/ADF2yADF10ADFagXGby3HYyWTTCkvA/P3095/Data-R.E.D.Server/Mol_MM/MM_sander_vs_pyred_SinglePoint/Mol_mm4-c1.log
i.e. for NSTEP = 1
BOND = 0.4776 ANGLE = 0.0827 DIHED = 0.0000
VDWAALS = 0.0000 EEL = 0.0000 HBOND = 0.0000
1-4 VDW = -0.0061 1-4 EEL = 7.2098 RESTRAINT = 0.0000
ANGLE energy should equal 0.0126 !!!
---
I tested our old sander 12 with job P2875; it looks like same problem...
---
and GAFF version 2.2.30 has still duplicates as discussed several
months ago...
The user gets such an email:
Identified force field in 'frcmod.user': GAFF version 2.2.30 - 2025
Duplicated FF parameters found in the 'frcmod.user' file:
Duplicated_improper_dihedral_angle_FF_parameter 'c-c2-c2-c3' found 2 times
Duplicated_improper_dihedral_angle_FF_parameter 'c-ca-ca-c3' found 2 times
Duplicated_improper_dihedral_angle_FF_parameter 'ca-n2-ca-n2' found 2 times
DUPLICATED FF PARAMETERS FOUND IN THE 'frcmod.user' INPUT FILE
CORRECT THE 'frcmod.user' FILE, AND REMOVE THESE UNWANTED FF PARAMETERS
Reordered FF parameters found in the 'frcmod.user' file:
Reordered_improper_dihedral_angle_FF_parameter found: ca-n2-ca-n2
<--> ca-n2-ca-n2
REVERSED/REORDERED FF PARAMETERS FOUND IN THE 'frcmod.user' INPUT FILE
CORRECT THE 'frcmod.user' FILE, AND REMOVE THESE UNWANTED FF PARAMETERS
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Jul 08 2026 - 22:30:02 PDT
