LYNX: A Multimodal Spatial Omics Framework

elhamazizi · x · 2026-07-12

LYNX combines cross-modal graph attention with interpretable Bayesian modeling to infer dynamic cell-cell communications directly from multimodal spatial data. The framework also scales from 2D tissue slices to emerging 3D multimodal datasets. On liver, breast cancer, and thymus data, LYNX identifies spatial multicellular programs and communication patterns that single-modal approaches struggle to recover.

Related event: LYNX: Multimodal Spatial Framework for Cell Communication(2 posts)→

Original post →

More from Research

Research channel →