Unsupervised analysis of whole transcriptome data from human pluripotent stem cells cardiac differentiation

Area of Science:

• Stem cell biology
• Cardiovascular research
• Bioinformatics

Background:

• Pluripotent stem cell (PSC) differentiation protocols vary, impacting cardiac cell development.
• Understanding gene expression differences is crucial for optimizing differentiation efficiency.

Purpose of the Study:

• To compare gene expression patterns across different PSC cardiac differentiation protocols.
• To identify novel gene expression markers predictive of differentiation efficiency.

Main Methods:

• Utilized unsupervised machine learning algorithms to analyze transcriptomic data.
• Compared gene expression profiles from cells cultured in 2D monolayers versus 3D aggregates.

Main Results:

• Visualized transcriptomic changes during differentiation, revealing similarities and differences between protocols.
• Corroborated existing findings and identified new gene expression patterns.
• Found a correlation between early non-mesodermal gene expression and lower cardiomyocyte differentiation efficiency.

Conclusions:

• Developed a versatile workflow for analyzing stem cell differentiation transcriptomes.
• Identified potential predictive markers for cardiomyocyte differentiation efficiency.
• The workflow holds promise for future clinical applications in cellular therapy.