Topic 464 Phase I Award: Network-Based Multimodal Data Analysis Cloud Software for Cancer Biomarker Discovery and Drug Development

Sostos LLC will deliver cloud-based multimodal data analysis software enabling users to discover biomarkers and develop new/repositioning drugs to improve cancer outcomes.

Current multimodal data analysis methods are computationally inefficient or ineffective for making actionable biological discoveries.

Our cutting-edge AI technology can efficiently model multi-omics networks in various patient sample types, leading to our discovery of cancer biomarkers with FDA “Novel Technology” status and potential new drugs. Our AI technology outperforms other methods in comparative studies.

Our goal is to establish a cloud-based AI software platform to accelerate biomarker discovery and drug development to improve personalized cancer diagnosis and prognosis, optimal treatment recommendations, and new therapeutic options. Our current software pipeline is implemented in a local computing environment.

To enhance the usability and interpretability of our AI technology and increase its market value, this Phase I SBIR project will test the feasibility of developing a cloud-based software prototype of our AI pipeline with a user-friendly graphical interface for visualization. This prototype will be tested with The Cancer Genome Atlas (TCGA) and Cancer Cell Line Encyclopedia (CCLE).

In Phase II, this prototype will scale up, include other data sources and customer input data, and ensure rigorous data security/safety/privacy.