Insilico CEO Says AI Cuts Time to Drug Candidate to About One Year in China

Insilico Medicine CEO Alex Zhavoronkov told Reuters that the company typically reaches a developmental drug candidate in 13 months, with a record of nine months, by combining AI with its research operations in China. He said the traditional approach usually takes about 4.5 years to reach the same stage. AI has helped Insilico generate 31 developmental candidates over six years, and its first AI-designed drug, Rentosertib, has advanced to Phase II clinical trials. The company's Shanghai laboratory has automated biological sampling and screening, while no experimental medicines developed by Insilico have yet been approved for sale.

AI-driven drug discovery timelines

• ▪Insilico Medicine CEO Alex Zhavoronkov said the company typically reaches a developmental drug candidate in about 13 months by combining AI with its China research ecosystem.
• ▪Novartis CEO Vas Narasimhan estimates it will take 8 to 10 years to fully realize the scale of artificial intelligence's impact on drug discovery timelines.
• ▪Insilico Medicine CEO Alex Zhavoronkov said the traditional approach usually takes about 4.5 years to reach a developmental drug candidate.
• ▪Insilico Medicine CEO Alex Zhavoronkov said the company's record time to a developmental drug candidate is nine months, while its typical timeline is 13 months.
• ▪Insilico Medicine has generated 31 drug development candidates using artificial intelligence over the past six years.

Insilico Medicine IPF candidate

• ▪Insilico Medicine's flagship small-molecule drug candidate, Rentosertib, was designed using generative artificial intelligence to treat idiopathic pulmonary fibrosis.
• ▪Insilico Medicine's first AI-designed drug, Rentosertib, has advanced to Phase II clinical trials.
• ▪Insilico Medicine utilized generative adversarial networks and reinforcement learning to identify the biological target and generate chemical structures for INS018_055 in 18 months.

Biological aging clocks

• ▪Quantitative biological aging clocks provide reproducible biomarkers that allow researchers to evaluate longevity interventions without running decades-long clinical endpoints.
• ▪Deep learning algorithms analyze multi-omic datasets, including DNA methylation, transcriptomic profiles, metabolomic signatures, and blood chemistry, to calculate biological age.

Longevity biotech financing models

• ▪Insilico Medicine has signed research and development partnerships with Eli Lilly, Takeda, and a potential $2.5 billion collaboration with Bora Pharmaceuticals.
• ▪Specialized artificial intelligence longevity startups secure non-dilutive capital through multi-billion-dollar upfront licensing deals and milestone-based co-development partnerships with multinational pharmaceutical corporations.
• ▪Longevity enterprises utilize a dual-track business model, discovering molecules for fundamental aging biology while validating them in near-term, disease-specific clinical indications.

Automated laboratory infrastructure

• ▪Tsinghua University Professor Yu Li said research and development costs had fallen 25% since the introduction of AI.
• ▪Researchers presented a materials-analysis system combining AI agents and a robotic arm to automate equipment operation and data analysis at a Huawei-hosted AI4S event.
• ▪Insilico Medicine's Shanghai research and development laboratory has automated the biological sampling and screening process.

6 sources