Fig. 4: Meta-analysis of gene expression across developmental T21 datasets.
From: Trisomy 21 Drives ADARB1 Overexpression and Premature RNA Recoding in the Developing Fetal Brain
A Volcano plot displaying transcriptome-wide differential expression results from a meta-analysis of ten independent T21 versus control RNA-seq studies. For each dataset, differential expression was estimated using a covariate-adjusted linear model with empirical Bayes moderation (two-sided). Dataset-specific log₂ fold-changes (x-axis) were synthesized using a random-effects meta-regression model to account for between-study heterogeneity; P values were adjusted using Benjamini–Hochberg FDR (y-axis). Genes are colored according to direction of dysregulation and chromosomal location, with chromosome 21 genes highlighted. The top ten T21-related over-expressed genes are labeled. B Average log₂ fold-change (y-axis) per chromosome, showing that chromosome 21 genes approximate the expected ~1.5-fold increase due to gene triplication. C Ranked plot of log₂ fold-changes for all significant chromosome 21 genes (FDR < 1%), arranged by genomic position (x-axis), illustrating selective rather than uniform dosage sensitivity. Eight well-known T21-related genes and ADARB1 are highlighted. D–G Forest plots display standardized mean differences in expression across ten datasets for four high-confidence dosage-sensitive genes: ADARB1, DYRK1A, IFNAR1, and IFNAR2. Forest plots show dataset-specific effect sizes and 95% confidence intervals, with summary estimates from a random-effects model (meta-regression framework). Summary diamonds (pink) denote the combined effect estimates derived from a random-effects meta-analysis, with axes scaled to the minimum and maximum confidence intervals for each gene to facilitate comparison.
