CellBRIDGE: Learning Cellular Trajectories via Interaction-Aware Alignment

2026-05-28

AI Summary

This study addresses the challenge of inferring cellular dynamics from destructive single-cell RNA sequencing snapshots, a task further complicated by the frequent neglect of ligand–receptor-mediated intercellular communication in existing trajectory inference methods. The authors propose an interaction-aware optimal transport model that explicitly incorporates typed, directed ligand–receptor interactions as part of the transport cost, jointly leveraging gene expression profiles to align cells across time points and reconstruct fate trajectories. This approach not only achieves higher trajectory inference accuracy—outperforming baseline methods that rely solely on gene expression features on both synthetic and real datasets—but also enables interpretable in silico perturbations to simulate signaling pathway inhibition. Applied to lung cancer data, the model reveals how specific intercellular communications regulate cell fate decisions.

Abstract

Inferring dynamics from population snapshots is a fundamental challenge in machine learning and biology. In scRNA-sequencing (scRNA-seq), destructive measurements preclude direct tracking of individual cells across time, making trajectory inference underdetermined. Optimal Transport (OT) provides a principled framework for snapshot alignment, but a long-standing modeling question is which cost functions yield biologically meaningful couplings. Standard OT approaches rely on gene-expression distances, implicitly treating cells as independent points and neglecting structured cell-cell communication mediated by ligand-receptor signaling. We introduce CellBRIDGE (Cell-Based Regularized Interaction-Driven Gene Expression), which augments feature-based OT with a directed, typed interaction cost derived from ligand-receptor activity. By explicitly modeling cell-cell communication, CellBRIDGE improves cross-snapshot couplings and downstream trajectory estimates across synthetic and real scRNA-seq datasets relative to feature-only baselines. Notably, CellBRIDGE enables mechanistically interpretable in silico perturbations: on lung cancer data, silencing specific ligand-receptor pairs induces trajectory shifts that recapitulate expected effects of targeted pathway inhibition.

Problem

• trajectory inference
• single-cell RNA-seq
• optimal transport
• cell-cell communication
• ligand-receptor signaling

Innovation

• Optimal Transport
• cell-cell communication
• ligand-receptor interaction
• trajectory inference
• scRNA-seq