Psi4 QM Conformer Ensemble (Part I of CSP Protocol)

Category Paths

Follow one of these paths in the Orion user interface, to find the floe.

  • Solution-based/Formulation

  • Role-based/Formulation Scientist

  • Product-based/Crystal Structure Prediction

  • Task-based/Crystal Structure Prediction


Generate all low-energy conformers of the input molecule with optional sub-sampling. This is the first step in OpenEye’s Crystal Structure Prediction protocol and uses the levels of theory and settings found to be the most useful for that application.

First, A large ensemble of potential conformations are generated using OpenEye toolkits, through exhaustive sampling of ring conformations, torsions, and hydrogens. To customize torsions rules used for this conformer generation, use the rules output from the Psi4 Combined Tautomer and Torsion Sampling Conformer Floe.

Next, the generated conformers are optimized using Psi4 at the HF-3c level of theory with all torsions constrained. The conformers are filtered to remove exact duplicates. The conformer energies are corrected to a higher level of theory (B3LYP-D3MBJ/6-31G*) through a single-point energy calculation. Last, high-strain conformations are removed at the specified energy cutoff.

This Floe assumes that all input molecules are a part of the same ensemble. Since conformers are generated at the beginning, there is no reason to provide more than one input molecule. However, this Floe will also work with multiple tautomers or stereo-isomers of the same molecule, if they are provided as input. The energy window is applied across all tautomers.

If an ‘Input Reference Conformers’ dataset is provided, then only conformers within the ‘Subsampling RMSD Threshold’ are kept after the conformer generation step. This step can be used to enable the multi-stage crystal structure prediction protocol if the unique conformers from the Polymorph Search with IEFF Crystal Force Field (Part II of CSP Protocol: Generation and Filtering) Floe are used as the reference input here. If sub-sampling, make sure to set the ‘RMSD Threshold For Omega Conformer Generation’ to a higher resolution than the sub-sampling threshold.

Promoted Parameters

Title in user interface (promoted name)

  • Omega conformations (confgen_data_out) type: dataset_out: Output dataset to write to
    Default: confs
  • Failures (failure_out) type: dataset_out: Output dataset to write to
    Default: failure
  • Input molecule (SMILES or conf ensemble) (in) type: data_source: The dataset(s) to read records from
  • Input Reference Conformers (in_ref_confs) type: data_source: Optional dataset with reference conformers. When no input is provided, all generated conformers are kept, otherwise, only those similar to reference conformers are kept in the workflow.
  • Quantum conformational ensemble (psi4_gopt_data_out) type: dataset_out: Output dataset to write to
    Default: confs_psi4_gopt
  • Quantum Geometries (All) (qm_geometries_data_out) type: dataset_out: Output dataset to write to
    Default: all_qm_geometries