Key points from article :
A Paris-based AI startup Bioptimus has secured $41 million in funding to accelerate the development of a universal AI foundation model for biology, bringing its total funding to $76 million. This model aims to address the fragmentation in biological research, where data is often studied in isolation, such as DNA, proteins, cells, or tissues. By integrating data across multiple scales and modalities, the model aspires to provide a unified understanding of biological systems, enabling researchers to analyse biology in its natural complexity.
The company has already made progress by launching one of the largest foundation models for pathology, which has demonstrated exceptional performance in predicting gene expression and accurately classifying certain diseases. Independent evaluations validated the model’s capabilities, showcasing its potential to outperform existing tools in these areas. These advancements are expected to contribute to better understanding disease mechanisms and improving precision in therapeutic design.
The next phase of the project involves developing a multi-scale, multi-modal foundation model that can simulate biological processes on a large scale. This tool will empower researchers to predict disease outcomes, understand treatment responses, and create therapies with high precision. By learning directly from raw data spanning molecules to entire organisms, this approach aims to tackle challenges across various industries, including pharmaceuticals and biotechnology.
The recent funding will support the refinement of existing models, the integration of more diverse datasets, and the formation of partnerships with pharmaceutical and biotech companies. The company’s mission is to create a transformative tool for biological research, enabling innovative solutions to complex problems and advancing industries reliant on a deeper understanding of biology. By leveraging cutting-edge AI, this initiative aims to break down research silos and unlock the full potential of integrated biological data for practical applications.
