top of page

Biomy, Inc. Selected for METI and NEDO’s "Post-5G Information and Communication Systems Infrastructure Enhancement R&D Project / Development of Competitive Generative AI Foundation Models"

  • 6 分前
  • 読了時間: 3分

– Developing the "Virtual Cell" Multimodal Spatial AI Foundation Model –



Biomy, Inc. (hereinafter "Biomy") announces that it has been selected for the "Post-5G Information and Communication Systems Infrastructure Enhancement R&D Project / Development of Competitive Generative AI Foundation Models," promoted by the Ministry of Economy, Trade and Industry (METI) and the New Energy and Industrial Technology Development Organization (NEDO). Biomy's adopted project is titled "Development of the Virtual Cell Foundation Model: A Multimodal Spatial AI to Establish Strategic Autonomy in Next-Generation Drug Deveopment."


GENIAC (Generative AI Accelerator Challenge) is a project aimed at strengthening generative AI development capabilities within Japan. Primarily through the provision of computational resources, it seeks to accelerate the development of competitive generative AI foundation models. This marks the fourth round of public selection since the program's launch in February 2024.


With the support of this project, Biomy will engage in the research and development of "Virtual Cell," a foundation model capable of computationally reconstructing the human tumor microenvironment. Through this, we aim to significantly shorten the timeline for new drug development and dramatically increase the probability of success (PoS) in clinical trials by predicting novel target molecules, stratifying patients for trials, and identifying optimal combination therapies in advance.


The difficulty of accurately predicting a drug's effects on the complex tumor microenvironment, coupled with behavioral differences between animals and humans, has been a major factor in clinical trial failures. Leveraging this project, Biomy will tackle these challenges by developing a model that extracts global spatial features from a single tissue section and learns the complex relationships between molecules and morphology.


Specifically, through joint research with the Japanese Foundation for Cancer Research (JFCR), we will construct a multimodal spatial dataset that integrates spatial transcriptome, protein (multiplex immunofluorescence imaging), and H&E image information from cancer pathology slides. Furthermore, with technical cooperation from the University of Tokyo, which possesses extensive expertise in pathology AI development, we will train and construct a foundation model capable of predicting the spatial distribution of gene and protein expression across an entire tissue under virtual intervention conditions.


For the constructed model, we will form a working group with the Japan Biological Informatics Consortium (JBIC) to conduct objective evaluations and verify its practical utility in the field through industry-academia-government collaboration. Additionally, data construction protocols and infrastructure operational know-how for large-scale distributed training obtained during development will be published via tech blogs and white papers. Upon successful demonstration, non-exclusive licenses will be provided, thereby contributing to the establishment of next-generation AI-driven drug development.


Comment from Teppei Konishi, President and CEO of Biomy, Inc. "We are deeply grateful for the robust support from METI and NEDO, as well as their high evaluation of our vision for personalized medicine using AI and our R&D plans for the 'Virtual Cell' multimodal spatial AI foundation model. To overcome the difficulties in predicting drug efficacy caused by the complex tumor microenvironment and to reduce the risk of clinical trial failures, our entire team will diligently advance this development alongside our strong partners, including the Japanese Foundation for Cancer Research, JBIC, and the University of Tokyo."


About Biomy, Inc. Biomy, Inc. is a technology company focused on advancing personalized medicine through artificial intelligence. We develop AI models and a surrounding platform to analyze complex biological data, including spatial biology and pathology imaging data, aiming to enable deep biological insights and support the development of novel therapeutics. For more information, please visit: https://www.biomy-tech.com/

 
 
bottom of page