BrainEXP-NPD

BrainEXP-NPD aggregates and standardizes human brain transcriptomic profiles for schizophrenia (SCZ), bipolar disorder (BD), major depressive disorder (MDD), autism spectrum disorder (ASD), Alzheimer’s disease (AD), and Parkinson’s disease (PD) to enable comparative analyses across these neuropsychiatric disorders.

Key Features:

• Extensive Data Collection: Integrates 48 brain transcriptome datasets comprising 8,317 samples across multiple brain regions, both sexes, and an age range from embryonic stages to seniors, including cases of SCZ, BD, MDD, ASD, AD, PD, and unaffected controls.
• Standardization and Quality Control: Performs data standardization, quality control, filtering, and removal of known and unknown covariates to enhance data consistency and reproducibility.
• Meta- and Mega-Analyses: Implements meta- and mega-analyses that include gene differential expression and gene co-expression network analyses.
• Differential Expression Thresholding: Identifies significantly differentially expressed genes using a significance threshold of FDR q < 0.05 across brain regions and disorders.
• Disease- and Region-Specific Patterns: Reports disease- and region-specific transcriptomic similarities and differences, for example maximal SCZ–BD similarity in the striatum, SCZ–MDD similarity in the hippocampus, and negative correlations between PD and SCZ/BD/MDD in cerebellum and striatum.

Scientific Applications:

• Cross-Disease Comparisons: Facilitates comparisons among SCZ, BD, MDD, ASD, AD, and PD to identify shared and distinct molecular mechanisms.
• Gene Function Studies: Supports investigation of gene functions and biological pathways using detailed brain transcriptomic profiles.
• Biomarker Discovery: Enables identification of candidate biomarkers and co-expression modules relevant for diagnosis, prognosis, and therapeutic targeting.
• Pathophysiological Insights: Allows examination of region-specific gene expression patterns to inform disease pathophysiology.

Methodology:

Computational steps explicitly include compiling 48 brain transcriptome datasets (8,317 samples), data standardization, quality control and filtering, removal of known and unknown covariates, followed by meta- and mega-analyses comprising gene differential expression and gene co-expression network analyses with significance assessed at FDR q < 0.05.