A cross-platform meta-analysis of human brain transcriptomes identifies core and tissue-specific molecular signatures in alcohol use disorder.

Alcohol use disorder is associated with widespread molecular alterations in the human brain, yet transcriptomic findings remain heterogeneous across studies and tissues. To address this, we performed a cross-platform meta-analysis integrating publicly available RNA sequencing and microarray datasets from multiple human brain regions. Using a random-effects framework with study-level clustering, we identified a set of reproducible gene expression changes that were consistent across independent datasets and robust to sensitivity analyses. The resulting core signature was characterized by coordinated alterations in pathways related to neuronal signaling and cellular architecture and also immune signaling, cellular metabolism, mitochondrial function, and protein homeostasis. Tissue-level analyses further revealed both shared and region-specific patterns, including context-dependent differences in gene expression direction across brain areas. The identified molecular signature demonstrated reproducible classification performance in independent datasets, including peripheral blood, suggesting that part of the identified signal may extend beyond brain tissue. Overall, these findings support the presence of a consistent and biologically structured transcriptomic profile associated with alcohol use disorder and provide a framework for future studies aimed at clarifying its molecular basis and evaluating its possible translational relevance.