QuantumXCT: Learning Interaction-Induced State Transformation in Cell-Cell Communication via Quantum Entanglement and...

Area of Science:

• Computational Biology
• Quantum Machine Learning
• Single-Cell Genomics

Background:

• Current cell-cell communication (CCC) inference relies on ligand-receptor databases, limiting discovery to known interactions and co-expression patterns.
• Existing methods fail to capture system-level signaling effects on cellular states.

Purpose of the Study:

• To introduce QuantumXCT, a hybrid quantum-classical generative framework for inferring CCC by learning interaction-induced state transformations.
• To enable de novo discovery of communication programs in complex biological systems.

Main Methods:

• QuantumXCT encodes transcriptomic profiles into a Hilbert space.
• Parameterized quantum circuits learn unitary transformations mapping non-interacting to interacting cellular states.
• The framework models CCC as learning state transformations between cellular distributions.

Main Results:

• QuantumXCT accurately recovers complex regulatory dependencies and feedback structures from synthetic and real-world data.
• Identified key communication hubs, including the PDGFB-PDGFRB-STAT3 axis in ovarian cancer-fibroblast co-cultures.
• The quantum circuit's topology provides interpretable interaction networks and quantifies influence.

Conclusions:

• QuantumXCT offers a generative framework for modeling intercellular communication by learning data-driven state transformations.
• This approach overcomes limitations of static interaction lookup, establishing a new paradigm for CCC discovery.
• Highlights the potential of quantum machine learning in advancing single-cell biology research.