SPARC Reveals Distinct Osteoclastogenesis-Associated Malignant States in Metastatic Osteosarcoma

SPARC reconstructs directionally constrained malignant-cell trajectories across paired conditions, improving recovery of metastatic precursors and complete progression routes in simulations and lineage-tracing datasets.

• Why it matters: Static primary-versus-metastasis comparisons can miss how transcriptional programs change along progression routes, and existing trajectory tools often fail to jointly model within-condition ordering and cross-condition transitions.
• What they did: The authors introduce SPARC, a unified graph-based optimization framework that evaluates within-condition progression and cross-condition correspondences under a common connectivity formulation, then test it in simulations, time-course settings, and real lineage-tracing data.
• The result: In a paired mouse scRNA-seq model of primary osteosarcoma and spontaneous lung metastases, metastatic tumor states differ in their association with osteoclast-like myeloid populations and supportive signals, suggesting some states may preferentially promote osteoclastogenesis.

AI summary of the abstract · may contain errors · not medical advice · read the paper.