As a newly appointed SciLifeLab Fellow, my research program in computational biology aims to investigate human stem cells and fibroblasts across adult healthy and diseased tissue samples, focussing on understanding stem-cell associated tissue niches involving epithelium, vasculature, fibroblasts, and infiltrating immune cells. Based on single-cell atlases of scRNA-seq, scATAC-seq, multi-ome, along with emerging spatial transcriptomics data, it is increasingly feasible to discover tissue structures, cell-cell interactions, and gene regulation driving advances in the molecular biology of cells in health and disease. In addition, genome-wide association studies have provided insights into hereditary variants of complex diseases such as heart diseases, inflammatory disorders, and cancer. As such, we plan to investigate hypotheses surrounding how genetic disease risk may be explained by fine-mapping genetic variation to tissue structures involving stem cells and fibroblasts. However, computational methods for such multimodal analysis are presently limited, impeding effective scientific progress in several areas. To address these research gaps, I plan to develop innovative computational methods based on data from clinical and experimental collaborators, publicly available resources, and synthetic data, advancing the capability for: 1. Modelling gene regulation at base pair resolution by transcription factor-DNA occupancy and protein complexes. 2. Discovering tissue structures as composites of cells across multiplex imaging and spatial transcriptomics platforms. 3. Retrospective lineage tracing with DNA barcoding and somatic mutations. 4. Cross-organ comparative analysis of single-cell gene programs associated with stem cell properties. 5. Identification of ligands for decorating lipid nanoparticles to target stem cells with RNA therapeutics. By solving timely computational problems in these fields, we argue that this approach provides a substantial strategic advantage for making fundamental and translational research discoveries pertaining to stem cells for both my own laboratory and the broader research ecosystem of SciLifeLab. There is substantial potential for discovering stem cell and fibroblast-based mechanisms by innovative application of machine learning to train probabilistic or foundational models based on single-cell atlas data from the Human Cell Atlas and on data acquired through collaboration. The research will be focussed on 3 biological contexts of particular interest, which are positioned at the interface of the PIs expertise in cardiovascular and cancer biology: cancer immune evasion and mutagenesis, angiogenesis in cancer metastasis, and cardiac fibroblasts.