Quantum

QUBO-Based Conformer Selection with Quantum-Classical Hybrid Optimization for Multiconformer Ligand Modeling

Read full article at source ↗

Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 22/10/2026, 04:13:25 UTC.

Briefings

A quantum-classical hybrid solver using QUBO optimization was tested against an established MIQP approach for selecting ligand conformations from X-ray data. The study evaluated 83 paired ligand entries, finding that the quantum method produced higher ligand-region real-space correlation coefficients in 62 out of 83 cases and lower occupancy-weighted torsional strain in 66 cases. The study concludes that the formulation of conformer selection can alter which ligand states are represented from the same experimental data.

Key points

• Quantum-classical hybrid optimization using QUBO produced higher ligand-region real-space correlation coefficients in 62 out of 83 evaluated ligand entries compared to the MIQP-based qFit-ligand workflow.
• Quantum-classical hybrid optimization using QUBO produced lower occupancy-weighted torsional strain in 66 out of 83 evaluated ligand entries compared to the MIQP-based qFit-ligand workflow.

Context

This research addresses the challenge of multiconformer modeling in drug discovery, where ligands adopt multiple conformations within the same binding site that interact differently with the protein. The study compares a new QUBO-based approach against an established MIQP-based qFit-ligand workflow to determine which conformations to retain and assign occupancies.

Limitations

The study is limited to 83 paired X-ray ligand entries and uses a composite-omit map generated from the common input structure. Results are based on a preprint (arXiv:2610.29058v1) and may not yet be fully validated in experimental settings. The study does not provide causal claims about the physical mechanism of improved agreement, only that the formulation changed which ligand states were represented.

Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 22/10/2026, 04:46:02 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.

View archived briefing on IPFS
bafybeif6itacfu6vpert7qwft2c7morgpubsg6wettzfjmqxrj64d5aely