The addition of capability to import SCF wavefunctions (HF and DFT) from other QM packages (e.g. Orca, Gamess, Firefly, NWChem) would be a great feature to have.
Perhaps it could be done via MOLDEN files.
Importing SCF wavefunctions from other QM packages would be complicated. For example, the ordering of basis functions within one shell and the sign conventions are different from package to package.
Every QM package I listed either uses pure gaussians or allows the user to ensure pure gassians are used. The order of the pure gaussians has been standardized for many years. Since Q-Chem allows the use of user input basis functions, that look suspiciously like those of GAMESS, it’s not clear why there should be any issue.
Since the wavefunction is invariant with regard to sign convention, that should not be a problem.
Using the MOLDEN file as the transfer agent would obviate the concerns stated.
@jherbert Warning, please keep conversation at a professional level and refrain from personal attacks.
@jimkress I apologize for the treatment you have received here. It is strictly contrary to Q-Chem’s company and community culture. You are welcome to continue posting here or contact us at support@q-chem.com
To address the original question, there are indeed notorious issues related to inconsistent definition of atomic basis functions, including shell ordering, basis function component ordering (both spherical and Cartesian), and basis function normalization. If the MOLDEN standard unambiguously specifies those things, and I don’t know if it does, it would indeed be a good candidate for communicating MO information across different packages.
That said, could you provide more context for your request? Having a general capability is nice, but what are the circumstances that require you to use MOs from another package?
Thank you for your response.
Q-Chem offers numerous nice features that we will be using in our research on Cancer, specifically Ewing Sarcoma. That’s why our 501c3 Nonprofit spent thousands of dollars to get access to the program . We usually do not spend large amounts of money for software and did so in this case because we felt the feature set of Q-Chem was a good fit for the (very large and complex) systems we are studying.
However, there are times when other packages provide features that are not in Q-Chem, for example TRAH-SCF for guaranteed convergence in HF and DFT systems. I have already provided information on that to the Q-Chem developers.
So, for example, at the moment I would like to be able to import converged MOs from Orca (of which I have many) into starting or property orbitals for Q-Chem. That’s just the first example that comes to mind.