Elix, LINC Launch First AI Drug Platform with Fed Learning

Elix, Inc. and the Life Intelligence Consortium (LINC) have launched the world’s first commercial AI drug discovery platform powered by federated learning, a technique enabling collaborative AI model development without sharing proprietary data. The platform, Elix Discovery™, integrates models trained on data from 16 pharmaceutical companies, marking a major milestone in overcoming one of AI drug discovery’s greatest challenges: access to large, high-quality datasets.

Developed in partnership with Kyoto University and through Japan’s AMED-funded DAIIA initiative, the system leverages Elix’s federated learning library, kMoL, to ensure secure, decentralized training. The resulting AI models support predictions for on/off-target effects, ADMET properties, and molecular generation, offering pharmaceutical companies enhanced accuracy while maintaining confidentiality.

Also Read: Verified Clinical Trials launches in Japan to curb duplicates

This commercialization marks a unique shift from a government-funded research project to actual industry use. The DAIIA participants first adopted the platform. It should gain more users as more companies join. This will add data and enhance model performance. The initiative marks a big change in drug discovery. It shows Japan’s leadership in collaborative AI. This also sets a new standard for data-sharing in the pharmaceutical industry.