Fig. 5: Consensus gene expression profiles related to dynamic variation of regional cortical thickness. | Nature Communications

Fig. 5: Consensus gene expression profiles related to dynamic variation of regional cortical thickness.

From: Brain-wide and cell-specific transcriptomic insights into MRI-derived cortical morphology in macaque monkeys

Fig. 5: Consensus gene expression profiles related to dynamic variation of regional cortical thickness.The alternative text for this image may have been generated using AI.

a Cortical maps of regional CT at different ages, and corresponding weighted gene expression maps of regional PLS1 scores (weighted sum of 23,613 gene expression scores). The bottom panel shows scatterplots of regional CT (x-axis) versus regional PLS1 scores (y-axis) that are associated with a list of genes with rank denoted by the correlation between the first PLS component and respective gene expression maps. The blue line is the fitted line and gray shaded area indicates the 95% confidence intervals. p values are provided as a Source Data. CT, cortical thickness; PLS1, the first partial least squares component; L, lateral view; M, medial view. b UpSet plot showing intersections of PLS1 genes with normalized weights Z score > 3 or < −3 from age 2 to 8. The bar encoding the size of genes intersecting in all age groups is colored in blue. The similarity matrix shows the relationship between PLS1 scores of intersecting genes between different age groups. c Consensus CT-related PLS1 genes and corresponding functional annotation. Left: Venn diagram showing the overlap between identified PLS1 genes (2269 genes) and CT-correlated genes (1284 genes). Right: Genes that are strongly positively weighted on PLS1 are in red, while those negatively weighted ones are in blue. Loadings reflect the shared variance between the original variables and the PLS scores, which represents the degree of contribution of each variable to the latent variable identified by PLS. d Scatterplots of normalized expression of typical positively/negatively weighted PLS1 genes (e.g., L1CAM and UGT8, x-axis) versus averaged regional CT (y-axis) over major structures. Cortical maps illustrate that L1CAM exhibits relatively higher expression in regions with larger CT (Pearson’s R = 0.35, p = 0.00048), whereas UGT8 exhibits relatively lower expression in those regions (Pearson’s R = −0.37, p = 0.00017). e The number of the 1005 consensus CT-related PLS1 genes overlapped with those linked to CT in humans. R in a, d represent the Pearson correlation coefficient. p values in a, d are determined based on a two-sided test and not corrected for multiple comparisons. Each point in a, d stands for a cortical region (n = 97). Source data are provided as a Source Data file.

Back to article page