Pyxir®, our drug discovery platform, provides end-to-end solutions in preclinical drug discovery. The Pyxir® Platform has integrated advanced artificial intelligence technologies and domain knowledge from computational chemistry, medicinal chemistry and biology.
Evolved from tools and experience of medicinal chemistry, computational chemistry and biology, the Pyxir® platform provides end-to-end solutions in preclinical drug discovery such as molecular generation, property prediction, virtual screening and optimization. Using hundreds of models and world leading algorithms, the Pyxir® platform inspires drug discovery from multiple dimensions.
The Pyxir® Platform integrates domain knowledge of AI, computational chemistry, medicinal chemistry and biology to build a systematic project to meet challenges from research and development.
Galixir leverages cutting-edge AI algorithms to find candidate molecules with high potency, good druggability and novel structures. Our self-developed molecular generation and design models have stood out in diversity and novelty.
The requirements for target selectivity are attained by building AI models and using computational chemistry tools to capture subtle differences in targets and binding sites.
Based on multi-dimensional data, our ADMET property prediction models are used in virtual high-throughput screening of candidate molecules.
Galixir has already published nearly 20 papers in top journals and conference proceedings covering subjects including molecular generation, property prediction and virtual screening and so forth. And Galixir leads the world in drug design with optimizing computational models.
Project-specific data is trained using federated-learning algorithm to ensure performance as well as independence. Meanwhile, based on the project information isolation mechanism and authority management system embedded in Pyxir®, confidentiality is ensured in data transmission and analysis.
Galixir is the first company that applies Meta-learning methodology in lead optimization tasks, which enables us to achieve better prediction performance under the circumstance of data insufficiency in First-in-class projects.
Through collaboration with our strategic partners, we are developing pipelines on therapeutic areas including central nervous system diseases, autoimmune diseases, tumors, respiratory diseases. We have made good progress in both inhibitor development and agonist development. With rich experience in machine learning application, our AI engineering team is able to transfer our model advantages into project productivity.