Burkitt lymphoma (BL) carries multiple oncogenic drivers yet arises predominantly in children, whose normal cells harbour few age-related mutations. To investigate this paradox, we sought to define the sequence and timing of mutational drivers underlying BL development. Here, we analysed single-cell whole-genome sequencing (WGS) data of 250 paired normal and malignant B-cells from BL patients and integrated this data with 21 bulk WGS samples and an existing dataset of 157 clonally expanded B-cells from healthy individuals. Phylogenetic reconstruction revealed an accelerated accumulation of mutations following the emergence of the most recent common ancestor, leading to the early establishment of extensive genetic intra-tumoral heterogeneity. We further provide evidence that convergent evolution shapes this diversity at both the point mutation- and copy number variation-levels. This in-depth characterisation further establishes BL as a paradigmatic model of tumorigenesis with implications for therapeutic intervention.
Magnification-Aware Distillation (MAD): A Self-Supervised Framework for Unified Representation Learning in Gigapixel Whole-Slide Images
arXiv:2512.14796v1 Announce Type: cross Abstract: Whole-slide images (WSIs) contain tissue information distributed across multiple magnification levels, yet most self-supervised methods treat these scales as independent


