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.
01Generate ensemble

Enumerate reproducible 3D conformers and record random seeds.

02Calculate descriptors

MMFF94 energy, radius of gyration, and asphericity.

03Encode QUBO

Expose variables, coefficients, constraints, and objective assumptions.

04Compare baselines

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

MethodQuestionRequired output
Lowest MMFF energyDoes a simple baseline select the same conformer?Selected index and energy
Random selectionIs the objective better than chance?Seeded distribution
Classical SA/QUBODoes the encoding change selection quality?Objective, variables, iterations
Physical QPUIs 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.