QCElemental Molecule Format Error

Hello everyone, I’m parametrizing the ANP-PNP molecule with FFParam. Im on the step of fitting bonded parameters. When executing the QM Bond/Angle/Dihedral PES input files with Psi4 Im getting an error when the code reaches mol=psi4.geometry(psi4_xyz) and cant figure out how to fix it.

The error is: *** qcelemental.exceptions.MoleculeFormatError: Unprocessable Molecule remnants under [psi4, xyz, xyz+, psi4+]:
And a list of atoms with their xyz positions.

The input file is called anp_bondscan_o2bpb.py
The file with the coordinates is called anp_bondscan_o2bpb_xyz.py
Both are attached.

Just to get the full error:

Traceback (most recent call last):
  File "/home/leonardo/psi4conda/lib/python3.9/site-packages/qcelemental/molparse/from_string.py", line 256, in from_string
    molstr, molinit = parse_as_psi4_ish(molstr, unsettled=True)
  File "/home/leonardo/psi4conda/lib/python3.9/site-packages/qcelemental/molparse/from_string.py", line 217, in parse_as_psi4_ish
    raise MoleculeFormatError(f"""Unprocessable Molecule remnants under {dtype}:\n{molstr}""")
qcelemental.exceptions.MoleculeFormatError: Unprocessable Molecule remnants under [psi4, xyz, xyz+, psi4+]:
O1G    1.9190901294   -4.0300203396   -3.2098481477
H1G    2.0359745025   -3.4377604413   -3.9791216236
O2G    2.4604523004   -4.5343740115   -0.8514379430
O3G    4.4541321909   -3.9597571334   -2.4846493484
H2G    3.1617103015   -5.1048957042   -0.4855489058
O1B    3.1328022703   -1.3882767619   -3.9525439118
O2B    3.9587305230   -0.2520133938   -2.2181213521
N3B    2.7155945178   -2.1206394785   -1.5904805877
H1B    3.9483075438   -1.1348248846   -4.4277091966
H3B    2.2875693624   -1.9601969296   -0.6767630109
O1A    1.5095516458   -0.0600324027    0.0973132739
O2A   -0.3099242926    0.9775706016   -1.5470626055
O3A    1.7745966676    0.1260866533   -2.4810780493
O5'    1.8758876946    2.2344978183   -1.0787399106
C5'    2.1830312264    2.7969105619    0.2108347543
H2A   -0.7157451604    1.6201079914   -0.9191378920
H5'1    2.4626459458    1.9986300416    0.9042903140
H5'2    3.0378304773    3.4609769458    0.0579297566
C4'    0.9815482567    3.5752306570    0.7176510538
H4'    0.8045924304    4.4582777792    0.0922657164
O4'   -0.1904119785    2.7171372968    0.6197422537
C3'    1.1074429885    3.9472121582    2.1977205293
H3'1    2.1498527994    4.0121842660    2.5372973900
O3'    0.3821022744    5.1352003310    2.5195621163
H3'2    0.9036018164    5.9105772203    2.2494354293
C2'    0.3251875363    2.8311098779    2.8929754736
H2'1    0.9104522397    1.9052237769    2.9177816532
O2'   -0.0645664236    3.1234120771    4.2065577056
H2'2   -0.3323115791    4.0633183178    4.2015704521
C1'   -0.8352188630    2.6615582224    1.9102533776
H1'   -1.5626191210    3.4754867445    2.0179824837

Im using conda 4.10.3, psi-rt 1.5, psi 1.5+e9f4d6d, qcelemental 0.24.0, python 3.9.7
The old psi4 version is needed, due to psi4 output differences in new versions and FFParam parsers.

Thank you for your time!

anp_bondscan_o2bpb.py (1.6 KB)
anp_bondscan_o2bpb_xyz.py (30.3 KB)

Psi4 doesn’t recognize atom labels such as O1G. Current Psi4 recognizes either O1 or O_1G but not O1G. I don’t know what your version of Psi recognizes, but it’s almost certainly the same pattern. Primes are definitely not recognized.

The technical atom specification may be found here, at time of writing.

1 Like