Our workflow
Our ongoing work addresses a question that follows directly from these discoveries: mutations alone do not make disease. Because even healthy brains harbor small numbers of mutant cells, we ask which mosaic ecosystems actually transition to disease, and when and how that transition can be stopped. We are therefore reframing disease initiation as a threshold-crossing event determined by the interaction of cell of origin, microenvironment, and time — in which mutant cell populations, under various biological stress, progressively perturb surrounding cells and neural networks until cell states shift toward changes that are difficult to reverse.
Our strategy proceeds in four stages: (i) mapping the seeds by combining single-cell and spatial omics with ultra-rare variant detection; (ii) decoding transition mechanisms to identify the key regulators linking mutant cells, their ecosystem, and biological stress, including how mutant cells perturb surrounding neurons and neural network function (iii) building therapeutic platforms — ASO, siRNA, AAV, and epigenome editors — designed to control pathological state transitions rather than merely correct the underlying mutation; and (iv) advancing predictive and preventive medicine through cerebrospinal fluid biomarkers and AI-based risk prediction that enable intervention before clinical onset. This program is pursued together with Sovargen, a biotechnology company co-founded on the KAIST campus, and builds on our framework for RNA therapeutics targeting brain somatic mutations