Transcriptome Complexity Disentangled: A Regulatory Molecules Approach.

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Transcription factors (TFs) and microRNAs (miRNAs) are key regulators of gene expression and cellular processes.
• Understanding their role in predicting genome-wide expression is crucial for biological insights.

Purpose of the Study:

• To determine if a limited set of TFs and miRNAs can accurately predict genome-wide gene expression.
• To develop predictive models for gene expression using these regulatory molecules.

Main Methods:

• Analysis of 8895 cancer samples from The Cancer Genome Atlas across 31 cancer types.
• Unsupervised learning to identify miRNA and TF clusters, selecting medoids for model development.
• Development of Tissue-Agnostic and Tissue-Aware models to predict gene expression.

Main Results:

• Identified 28 miRNA and 28 TF clusters; medoids differentiated tissues with 92.8% accuracy.
• Tissue-Aware model achieved an R-squared of 0.70 by integrating tissue-specific data.
• Prediction accuracy using 56 molecules rivaled that of 1000 landmark genes, indicating a low-dimensional transcriptome structure.

Conclusions:

• A small subset of regulatory molecules can effectively represent the transcriptome, suggesting intrinsic low dimensionality.
• This approach enables cost-effective transcriptome assays and analysis of degraded samples.
• Provides insights into miRNA/TF regulatory roles versus alternative gene expression mechanisms.