Launch of a joint project with Qubigen to develop antiviral drugs using federated AI
Qubigen, a company specializing in AI-driven drug discovery, recently announced the launch of a new collaborative project focused on the decentralized (federated) AI-driven design of antiviral compounds. This joint project with the German Center for Infection Research (DZIF) and the Japanese Institute of Microbial Chemistry (IMC) focuses on the rapid development of novel, antiviral, small-molecule drugs against emerging infectious diseases.
Global collaboration among leading research institutions
This project is a collaboration between nonprofit organizations, academic institutions, and industry partners. The project combines their complementary expertise in antiviral research, computer-aided design, and translational drug development. Among other things, the project draws on relevant antiviral data provided as part of the joint "Nucleoside Booster" project between the DZIF and DNDi (Drugs for Neglected Diseases initiative), an international nonprofit research and development organization. In the Nucleoside Booster project, DZIF scientists from Heidelberg University Hospital and other DZIF member institutions collaborate with DNDi to identify chemical compounds with a broad spectrum of antiviral activity.
“Our collaboration with Qubigen enables us to leverage the data collected from the Nucleoside Booster Project to virtually train AI models on highly sensitive datasets, ensuring that the data remains in its original location and remains undisclosed,” says DZIF scientist PD Dr. Nadine Biedenkopf of Marburg University, who coordinates the collaboration between DZIF and Qubigen. "By securely consolidating insights from decentralized, global networks, we are laying the groundwork for the development of future medicines."
Another key partner in Qubigen and DZIF's antiviral initiative is the Institute of Microbial Chemistry (IMC) in Japan. This nonprofit research institute of the Foundation for Microbial Chemistry contributes antiviral data and scientific insights relevant to developing drugs against infectious diseases.
Secure Drug Design with Qubigen’s “Federated AI for Drug Design” (FedAIDD) Platform
A key component of the project is FedAIDD, a platform developed by Qubigen. It was designed to enable secure AI training across distributed datasets while preserving data privacy and intellectual property boundaries. FedAIDD brings together proprietary and public data sources within a collaborative framework, thereby supporting the generation of new molecular candidates without requiring participating organizations to disclose sensitive underlying data. The goal of this initiative is to generate novel small-molecule drug candidates with improved antiviral activity, greater synthetic feasibility, and enhanced drug-like properties. The aim is to shorten development times and advance next-generation drug discovery through secure, collaborative AI technologies. Possible avenues for further development and commercialization with an industry partner are also being explored.
Distributed (federated) search for promising antiviral compounds
To date, relevant datasets have been curated and uploaded to the secure FedAIDD platform. Qubigen has also initiated AI-driven design iterations to generate novel molecular candidates specifically tailored for antiviral activity. The goal is to identify promising new chemical compounds that can be prioritized for synthesis and biological validation. The most promising candidates will then advance to preclinical evaluation and further development. This multi-party collaboration highlights a key strength of the Qubigen’s FedAIDD platform: the combination of partner data, curated internal resources, and AI-driven drug discovery. Together, these elements support the design of novel molecules—a feat no single organization could achieve on its own. In this way, Qubigen’s technology can help accelerate drug design in areas with high unmet medical needs without compromising confidentiality.
Source: Press release of Qubigen
