Enumerate reproducible 3D conformers and record random seeds.
Research method · Classical optimization
QUBO conformer triage with explicit baselines.
Generate ligand conformers with RDKit, evaluate MMFF94 energy and shape descriptors, encode a selection problem as QUBO, and solve it with classical simulated annealing. Use the result to study representation and optimization—not to claim quantum docking or quantum advantage.
No quantum hardware is connected. No IBM Quantum, AWS Braket, D-Wave, Qiskit, or physical QPU execution occurs in the live workflow. The requested backend and the backend actually used must remain visible in every result.
MMFF94 energy, radius of gyration, and asphericity.
Expose variables, coefficients, constraints, and objective assumptions.
Evaluate whether the QUBO/SA selection adds value over simple minimum-energy and random baselines.
Run conformer experiment
Checking backend…No experiment run.
Required comparison
| Method | Question | Required output |
|---|---|---|
| Lowest MMFF energy | Does a simple baseline select the same conformer? | Selected index and energy |
| Random selection | Is the objective better than chance? | Seeded distribution |
| Classical SA/QUBO | Does the encoding change selection quality? | Objective, variables, iterations |
| Physical QPU | Is a connected quantum device better? | Not available today |
Promotion beyond Research requires reproducible benchmark evidence showing added value over simpler classical baselines on a relevant scientific task.