en.Wedoany.com Reported - ByteDance's artificial intelligence drug discovery unit has initiated a spin-off and is seeking independent financing.
After the spin-off, ByteDance will retain a controlling stake in the new company. The AI drug discovery unit's team, algorithms, technology platform, and existing pipeline assets will be transferred to the new entity, which will continue to receive computing power support from Volcano Engine. The unit, established in 2021 and led by Kai Liu, has approximately 50 core members. The team includes AI for Science (AI4S) talent, algorithm experts, and pharmaceutical experts, and has been engaged in foundational model research, platform development, and commercialization since its inception. ByteDance's previous team responsible for protein structure prediction models has also been integrated into this unit, with related algorithms and models already consolidated, and will continue advancing foundational model research in this field, though a few employees have left.
ByteDance has achieved multiple technological breakthroughs in AI drug discovery. In 2025, its AI4S team released molecular structure prediction models Protenix and Seedfold. In 2026, it launched Protenix-v1 and Protenix-v2, establishing open-source structure prediction capabilities for biomolecular systems such as proteins and ligands. In protein design and prediction, the team introduced tools like PXDesign for designing protein binders. ByteDance also launched Anew Labs, an AI-driven platform focused on real-world drug development. According to the Anew Labs website, the team has released research results including AnewSampling, AnewOmni, AnewFEP, AnewSynth, and scNext, covering areas such as protein-ligand dynamic structure prediction, full-atom molecular generation, free energy calculation, synthetic feasibility prediction, and virtual cells. The platform has also introduced early-stage drug pipelines.
In April, Anew Labs disclosed its IL-17 small molecule project for the first time at the American Association of Immunologists annual meeting. According to the disclosure, this is the world's first case of using a small molecule to block three IL-17 family dimers. IL-17 is a key pathway in autoimmune diseases such as psoriasis and ankylosing spondylitis. As ByteDance's AI drug discovery work matures, the company is attempting to transition it from research to industrial applications, having integrated relevant internal teams and preparing to test commercialization. The AI4S business has a long verification cycle and complex development processes; taking drug discovery as an example, it involves model research, wet lab experiments, and clinical validation, requiring specialized talent and an organizational model different from ByteDance's internet business.
The spin-off aims to establish an independent structure better suited for this business. ByteDance hopes this adjustment will help attract top talent and further integrate foundational model capabilities, algorithms, and pharmaceutical expertise. Healthcare data and clinical research services provider IQVIA expects global pharmaceutical spending to reach approximately $2.3 trillion by 2028. The market size is vast, but the high costs, long cycles, and high failure rates of new drug development have not fundamentally changed, driving demand for AI tools that can improve certain processes.
AI4S research is also accelerating, as exemplified by Google DeepMind's AlphaFold series of protein structure prediction models. AlphaFold 2 scaled up protein structure prediction massively, predicting approximately 200 million protein structures, while AlphaFold 3 expanded prediction to more complex interaction systems. Multimodal molecular generation models that have emerged in recent years address a more direct pharmaceutical challenge: drug design, indicating that AI drug discovery is gradually moving from research to industrial application.
ByteDance has been investing in AI4S for years. Around 2020, it began exploring areas such as AI drug discovery, molecular simulation, and computational biology, subsequently establishing teams covering first-principles calculations, quantum chemistry, molecular dynamics, materials simulation, and molecular generation for energy and pharmaceutical applications. After ByteDance established its Seed foundational model research team, AI4S became part of its frontier technology strategy. This marks ByteDance's first attempt to commercialize AI4S.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
