Persistent homology analysis of brain transcriptome data in autism.

Area of Science:

• Computational biology
• Neuroscience
• Genetics

Background:

• Persistent homology (PH) is effective for analyzing biological data with spatial or temporal components.
• Its application to gene expression data, particularly in neurological disorders, remains underexplored.

Purpose of the Study:

• To apply PH methods to gene expression data from post-mortem brain tissue of individuals with autism spectrum disorder (ASD) and controls.
• To investigate global topological and geometric properties of gene expression in ASD.

Main Methods:

• Persistent homology analysis was applied to gene expression data from cerebral cortex tissue.
• Key topological features, including the sum of death times of zero-dimensional components and Euler characteristic, were calculated.
• Analysis was performed on two independent datasets to ensure reproducibility.

Main Results:

• Significant differences in inter-sample geometric relationships were found between ASD and control groups.
• These differences suggest increased gene expression heterogeneity in individuals with ASD.
• No significant topological differences were observed at the gene-level point cloud analysis.

Conclusions:

• Persistent homology offers a novel framework for analyzing gene expression in complex genetic disorders like ASD.
• The findings indicate altered global gene expression patterns, specifically increased heterogeneity, in the autistic brain.
• The study highlights the potential of topological data analysis in understanding the neurobiology of ASD.