New drug development is a high-risk, high-return structure. From early research, massive expense and time are invested, but the failure rate is high, so only large corporations or advanced economies are taking the lead. In fact, 25%–30% of new drug development expense goes into early research and development (R&D), but the probability of entering the preclinical stage is only 0.05%.
To overcome this hurdle, pharmaceutical companies worldwide have adopted CADD (Computer-aided drug design, a technology that uses computer simulations to design and predict new drug candidates) to raise success rates and cut expense. A prime example is Schrodinger, the global No. 1 company used by 75% of the world's top pharmaceutical companies.
Korean Start - Up Atomatrix has thrown down the gauntlet in this market. Chief Executive Lee Eun-ho, who handled new drug development at SK Biopharm for 27 years and experienced firsthand the urgency of cutting research budgets, founded the company in May 2024. The company is led together with pharmaceutical experts including Executive Vice President Lee Sang-bae, who has worked in CADD for 29 years.
Lee said, "Schrodinger is designed around physics and computational chemistry experts, making it hard for new drug researchers to use directly, so the experiment success rate was only 10%," and noted, "Also, even though there is demand in the market for new drug development using the entire proteome, not just proteins with resolved structures, current CADD remains in a state where structural optimization has not been achieved."
Atomatrix set out to solve this efficiently and quickly, developing CADD to raise new drug success rates. Atomatrix developed two platforms, "BARon" and "Alopiper."
BARon is a binding prediction program in which the platform determines new drug candidate substances and drug target substances. It plays the same role as platforms including the existing Schrodinger but raised the new drug success rate from 10% to 50%. Lee explained, "First, we use AI protein structure prediction technology (Alphafold) to screen new drugs, and second, we use molecular dynamics-based new drug screening to raise the success rate."
Alopiper predicts the pharmacological signals that occur after a drug and a drug target bind. Lee explained, "Even if the binding of a drug and a drug target predicts the potential as a new drug, it is meaningful only if we can predict what actual efficacy occurs," adding, "Alopiper can predict a drug's toxicity and efficacy."
Atomatrix also focused on improving usability. Even without assigning a CADD expert to a research project, it provides a one-to-one optimized, customized module automation package so that researchers can directly use CADD to simulate new drug candidate substances.
The AI new drug development market is forecast at 200 billion won in 2030. The global market is projected at about 9 trillion won in 2029. Lee said, "Cases of big deals in new drugs are increasing at home and abroad," adding, "Riding this trend, we are expanding our business domestically and overseas." Atomatrix will establish a U.S. subsidiary in 2026 and enter the North American market. The 2029 revenue target is 20 billion won.
The market entry method is B2B customizing. Lee explained, "At the starting stage, we will enter the market by designing new drugs on behalf of corporations, and once clients feel the rise in their new drug success rates, we will move to selling optimized and automated software to customers."
In the mid to long term, the company will pursue a CADD SaaS (Software as a Service) business and co-develop new drug pipelines. Lee expressed ambition, saying, "After we secure recognition, we will launch a mass-market CADD service business like Schrodinger and take on high-value commercialization through co-development of new drug pipelines."