Identifying a ubiquitous gene expression variation pattern in the human brain

Chien-Ming Lo,Niall W Duncan

DOI: https://doi.org/10.1101/2024.07.21.604497

2024-07-22

Abstract:The availability of dense sampled gene expression data from across the human brain has allowed important investigations into fundamental brain principles. Correlated expression patterns of genes tied to microstructural properties have been studied for their relationships with diverse brain features. This work looks at the specificity of these relationships based on the sets of genes targeted. We find that the same spatial pattern emerges from any set with more than 180 genes in it. Looking at the association between this pattern and cortical myelination, as represented by a T1w/T2w map, we show that correlations between myelination and theoretically guided gene sets do not differ from random ones. This observation prompts a reevaluation of current methodologies and assumptions in the study of gene-brain associations. Additionally, our research highlights covariance characteristics within specific functional networks, underscoring their significant role in shaping the spatial transcription patterns in the brain. These findings contribute insights into the complex relationships governing brain organisation and gene expression patterns. They also highlight an important methodological factor influencing the validity of inferences made from gene expression covariance patterns.

Neuroscience