AI Drug Discovery Halves Costs, Boosts Success Rates

The drug candidate discovery platform 'CANDDIE', which combines artificial intelligence (AI) with molecular dynamics simulation, is showing results that cut costs by half and increase success rates compared to conventional methods.

Developed by startup AtoMatrix, the 'CANDDIE' platform simplifies the drug candidate discovery process by integrating AI technology and computational automation, with particular strengths in developing drugs targeting cell membrane proteins.

AtoMatrix CEO Lee Eun-ho stated, "We have never failed in the drug candidate discovery process with simulations that add molecular dynamics to AI," adding that "it is a platform that general drug developers can use without needing a CAD specialist."

CANDDIE's differentiating point is its prediction of dynamic molecular movements by incorporating molecular dynamics, with Lee explaining that it halves costs and reduces the development period to 4-6 months compared to existing discovery methods, while increasing success rates by more than fivefold.

AtoMatrix has achieved a 50% success rate for oral dual agonists targeting GLP-1R and GIPR, and a 65% success rate for immuno-oncology drugs. The company is currently conducting drug design research services for four companies, including HK inno.N and JW Pharmaceutical.