Spatiotemporal constrained RNA-protein heterogeneous network for protein complex identification

Area of Science:

• Computational Biology
• Systems Biology
• Bioinformatics

Background:

• Protein complex identification is vital for understanding cellular functions and diseases.
• Current methods struggle with noisy protein interaction data and ignore regulatory factors.
• Accurate protein complex identification requires accounting for spatiotemporal dynamics and biomolecular regulation.

Purpose of the Study:

• To develop a robust method for protein complex identification using heterogeneous RNA-protein networks.
• To address limitations of existing methods by incorporating spatiotemporal constraints and regulatory roles.
• To improve the accuracy and biological relevance of identified protein complexes.

Main Methods:

• Constructed a multiplex heterogeneous RNA-protein network incorporating spatiotemporal patterns.
• Employed a dual-view aggregator to integrate information from different network layers.
• Utilized contrastive learning to optimize protein embeddings based on spatiotemporal interactions.
• Applied a core-attachment strategy for protein complex identification, reweighting interactions based on embedding similarity.

Main Results:

• STRPCI effectively identifies protein complexes by considering spatiotemporal constraints and regulatory factors.
• The method demonstrates improved accuracy and biological significance compared to existing approaches on real-world PPI networks.
• STRPCI mitigates the impact of noisy data in protein interaction networks.

Conclusions:

• STRPCI offers a powerful new approach for protein complex identification.
• The method's ability to handle noisy data and incorporate regulatory factors enhances biological insights.
• STRPCI provides a valuable tool for researchers studying protein function, cellular processes, and disease mechanisms.