AI-assisted Analysis of Single Molecule Images for High-throughput Biomolecular Characterization

Cel

Project ATLAS revolutionizes single molecule analysis by building an AI-assisted platform to interpret multidimensional biomolecular images from scanning probe microscopy (SPM). This tackles two bottlenecks in biomolecular research: (i) the time-consuming manual interpretation of multidimensional SPM images, and (ii) the lack of singular digital environment, where agentic AI could assist users in interpreting image data. Project ATLAS aims to demonstrate AI-assisted multilayer image analysis and model building of biomolecules in a single, expandable, AI-integrated digital environment, capable of processing hundreds of single molecule images into their molecular structures and summarizing key insights from the dataset. The core project activities include building the software, integrating AI models to the software, deploying the software to beta users, developing an IP strategy, and formulating business plan. Project ATLAS main innovation lies in providing a one-stop environment for multilayer data analysis and computational chemistry (including molecular visualization/modeling) that can be directly acted upon by agentic AI models. Integrating agentic AI model into this digital laboratory is key to increase data throughput and lower adoption barriers of SPM-based biomolecular imaging technology for non-experts. The high-throughput single molecule analysis enabled by Project ATLAS holds significant potential across structural biology, AI-driven laboratory automation, and biotechnology sectors, which ultimately translate into cost reduction and acceleration of therapeutics development. Project ATLAS bridges cutting-edge science and technology with strategic commercialization plans, thus translating innovative research into a commercial product with significant impact on single molecule analysis markets that would reinforce European leadership in this emerging field.