Tuulia Vallius

This study identifies genomic subtypes and immune features of mucosal melanoma, linking CD163+ macrophages to immune resistance and suggesting TIM-3 blockade as a therapeutic target.

STITCH pairs size-independent tangent-PCA contour shape modes with expression factors by canonical correlation, recovering keratinocyte, fibroblast and melanoma shape-expression links in Xenium and CosMx data.

The paper shows that keratinocyte EMT-like transformation in melanoma reduces tumor invasion and improves survival through altered cell-cell interactions.

This study shows that immunogenic tumor mass dormancy, not cell-level quiescence, persists in residual melanoma lesions after immunotherapy, with active tumor cells and immune interactions.

This study shows CD40 agonist therapy reduces tumor growth in antigen presentation-deficient tumors by reprogramming monocytes and requiring IFN-gamma and CD8+ T cells for efficacy.

This paper shows that bleb-induced septin signaling supports survival of metastatic melanoma cells in vivo, and disrupting this pathway reduces metastasis.

The paper maps oxidative stress response in TNBC using single-cell and spatial data, identifying GCLC+ mesenchymal tumor cells linked to better patient outcomes.

The paper identifies melanoma differentiation and immune exhaustion as predictors of tumor progression and therapy response in in-transit melanoma patients through multiomic profiling.

The paper maps spatial patterns of tumor cell dedifferentiation and plasticity in melanoma by analyzing RNA and protein expression across 300 histological domains.

This study shows chemotherapy causes myeloid-driven spatial T-cell exhaustion in ovarian cancer by altering immune cell states and interactions, suggesting new immunotherapy approaches.

This study shows PRAME expression in melanoma is negatively regulated by TET2-mediated DNA hydroxymethylation, with reduced 5hmC levels correlating with PRAME upregulation.

SpatialCells automates the analysis of spatially resolved multiplexed single-cell data to characterize tumor microenvironments, enabling efficient feature extraction and downstream machine learning applications.

The paper introduces two methods to improve tissue segmentation in highly multiplexed images using real augmentations and nuclear envelope imaging, resulting in more accurate models.