LRLoop: a method to predict feedback loops in cell-cell communication.

Area of Science:

• Cellular biology
• Computational biology
• Bioinformatics

Background:

• Intercellular communication is crucial for multicellular organisms, coordinating biological processes.
• Feedback loops between cell types are vital signaling motifs in development, regeneration, and cancer.
• Existing computational methods often predict unidirectional cell-cell communication, failing to identify feedback loops.

Purpose of the Study:

• To introduce LRLoop, a new method for analyzing cell-cell communication via bidirectional ligand-receptor interactions.
• To assess LRLoop's performance against existing methods using bulk and single-cell gene expression datasets.
• To apply LRLoop to identify novel feedback loops in retinal development.

Main Methods:

• LRLoop identifies pairs of ligand-receptor interactions that are mutually responsive, forming a closed feedback loop.
• Method performance was evaluated using bulk datasets, assessing false positive rates.
• A novel strategy using between-tissue interactions as a false-positive indicator was developed for single-cell datasets.

Main Results:

• LRLoop significantly reduces the false positive rate compared to traditional methods on bulk datasets.
• LRLoop demonstrates a lower fraction of between-tissue interactions, indicating improved accuracy on single-cell data.
• Application to retinal development datasets revealed numerous novel bidirectional ligand-receptor interactions potentially regulating key developmental processes.

Conclusions:

• LRLoop is an effective method for identifying bidirectional cell-cell communication feedback loops.
• The method offers improved accuracy and reduced false positives in both bulk and single-cell analyses.
• LRLoop facilitates the discovery of novel signaling pathways involved in biological processes like development and regeneration.