Weekly Recap: 97% valid molecules in quantum AI and ¥40 million NEDO award

What happened

Shionogi and D-Wave said their quantum-assisted AI model produced 97% chemically valid molecules, versus 73% for a classical model, with more strong drug-likeness scores. Shionogi also won NEDO's Quality of Life category, securing ¥40 million for further drug-discovery research.

Confirmed

Global impact / market context

Higher valid-molecule rates could cut wasted lab work in early drug discovery, which means fewer costly experiments before promising compounds are found. The ¥40 million award funds further research, supporting Shionogi's spending on quantum computing tools without relying solely on its own cash.

Analyst inference

The comparison against a classical model is the key signal: it suggests quantum-assisted methods may improve on standard AI approaches, which means faster candidate screening. If repeated at larger scale, that could shift research spending toward quantum platforms and raise competition among drugmakers adopting them.

Analyst inference

What to watch

1. The reported result was 97% chemically valid molecules versus 73% for a classical model, with more strong drug-likeness scores. Drug-likeness scores estimate how likely a molecule is to become an oral medicine. Confirmed
2. Watch whether Shionogi and D-Wave publish peer-reviewed or larger-scale results, which would test whether the 97% validity rate holds beyond this initial comparison against the classical model. Proposed
3. Track how the ¥40 million NEDO award is spent and whether it funds new quantum hardware, staff, or partnerships. Such spending could signal whether drugmakers commit more capital to quantum-assisted discovery. Analyst inference

Evidence

• Original source link unavailable