Novel 4D Tensor Decomposition-Based Approach Integrating Tri-Omics Profiling Data Can Identify Functionally Relevant...

Area of Science:

• Molecular Biology
• Systems Biology
• Bioinformatics

Background:

• Gene expression regulation involves multiple layers beyond transcript abundance.
• Integrating transcriptome, translatome, and proteome data is crucial but lacks standard frameworks.
• Ribosome profiling and proteomics aid in distinguishing translation dynamics.

Purpose of the Study:

• To propose a novel unsupervised framework for tri-omics integration.
• To apply tensor decomposition for feature extraction across transcriptome, translatome, and proteome.
• To identify gene clusters associated with specific regulatory mechanisms like ribosome stacking and translational buffering.

Main Methods:

• Four-dimensional tensor decomposition (higher-order singular value decomposition) applied to tri-omics data.
• Analysis of transcriptome, Ribo-seq (translatome), and proteome profiles.
• Gene selection and enrichment analyses to interpret identified components.

Main Results:

• Identified components reflecting ribosome stacking and translational buffering.
• Selected 1781 genes linked to ribosome stacking and 227 to translational buffering.
• Enrichment analyses revealed pathway associations for these gene clusters, including translation, cell cycle, and stress responses.

Conclusions:

• Tensor decomposition provides a robust method for unsupervised tri-omics integration.
• The approach effectively extracts biologically interpretable components like ribosome stacking and translational buffering.
• This framework aids in identifying functionally relevant gene clusters from complex multi-omics datasets.