MultiScale AI for SingleCell-Based Precision Medicine

Cell-based precision medicine holds transformative potential for healthcare and requires a comprehensive understanding and integration of disease variability across multiple scales. Single-cell multi-omics offers unique insights by providing molecular profiles and predictive biomarkers that are crucial for understanding disease trajectories. Leveraging heterogeneous data through innovative AI breakthroughs is essential for unraveling multiscale disease processes and advancing personalized treatments.

Our consortium aims to bridge the gap between single-cell data and targeted therapies by addressing key methodological challenges. Single-cell data enables the resolution of cell type confounds and reveals disease heterogeneity, especially when integrated with imaging for spatial analysis. To address disease complexity, we will develop AI methods for handling multiscale data and integrating genomic information with other assays.

Our approach will account for challenges such as high dimensionality and data scarcity, using causal, logical, and stochastic modeling for heterogeneous data integration. We will develop network inference methods to characterize molecular mechanisms underlying clinical phenotypes, combining symbolic AI with machine learning to improve predictive model characterization. Identifying key molecular targets and predicting therapeutic impacts in clinical samples are primary objectives.

The central challenge of precision medicine lies in integrating variability across different levels of cell decision-making, requiring a paradigm shift toward viewing disease as a predictable and controllable stochastic state. Achieving predictive capability demands coupling intracellular dynamics with cell population dynamics and reconciling mathematical frameworks with biological knowledge across multiple scales. Our ultimate goal is to construct predictive, executable models similar to digital twins, enabling data-driven targeted treatment strategies based on controlling cell fate decisions.