0
0The paper outlines an innovative framework for mapping the spatial organization of cell surface proteins to discover tumour-associated proximity antigens. By combining high-throughput microenvironment mapping with graph-based machine learning, researchers can identify protein co-targets based on disease-specific membrane proximity rather than expression levels alone. This approach successfully isolated the EGFR-CDCP1 pair as a clinically actionable target combination that enhances the efficacy of multispecific cancer therapeutics. Ultimately, this methodology establishes membrane proximity as a critical principle for designing advanced precision treatments with improved tumour specificity and reduced toxicity.
References:
Scandore C, Malone C F, May C K, et al. Proximity-guided graph learning reveals tumour-associated proximity antigens[J]. Nature, 2026: 1-11.
