Population-scale cross-disorder atlas of the human prefrontal cortex at single-cell resolution

Abstract

Neurodegenerative diseases and serious mental illnesses often exhibit overlapping characteristics, highlighting the potential for shared underlying mechanisms. To facilitate a deeper understanding of these diseases and pave the way for more effective treatments, we have generated a population-scale multi-omics dataset consisting of genotype and single-nucleus transcriptome data from the prefrontal cortex of frozen human brain specimens. Encompassing over 6.3 million nuclei from 1,494 donors, our dataset represents a diverse range of neurodegenerative and serious mental illnesses, including Alzheimer's and Parkinson's diseases, schizophrenia, bipolar disorder and diffuse Lewy body dementia, as well as neurotypical controls. Our dataset offers a unique opportunity to study disease interactions, as 21% of donors had comorbid diagnoses of two or more major brain disorders. Additionally, it includes detailed phenotypic information on neuropsychiatric symptoms, such as apathy and weight loss, which commonly accompany Alzheimer's disease and related dementias. We have performed stringent preprocessing and quality controls, ensuring the reliability and usability of the data. As a commitment to fostering collaborative research, we provide this valuable resource as an online repository, enabling widespread analyses across the scientific community.

Fig. 1

Overview of the dataset collection process and key outputs of the study.

Fig. 2

Summary of demographics and clinical data of the PsychAD cohort. (a) Overlap of the PsychAD cohort with MSSM AMP-AD, CommonMind and RADC cohorts. (b) Correlations among AD-related phenotypes. This analysis includes donors with either sole AD diagnosis (without comorbidities) or control samples (free of any diagnosis). For the “cognitive impairment” phenotype, untransformed CDR values are used for MSSM donors. RADC donors are numerically scaled as follows: Mild Cognitive Impairment (MCI) = 0.75, clinical dementia = 3. (c) Distribution of the number of diagnoses per donor. Note that “Dementia” and “MCI” are not counted as separate diagnoses if the donor already has a neurodegenerative or neurological disease. Also, NPS are excluded from this comparison. (d) Analysis of the counts and intersections among the top 10 most frequently represented diagnoses plus controls, with a minimum intersection size for plotting set to 10. FTD: Frontotemporal dementia; ASCVD: Atherosclerotic cardiovascular disease; PD: Parkinson’s disease; BD: Bipolar disorder; Diabetes: Diabetes mellitus Type 1/2/unspecified; Vascular: Vascular dementia; DLBD: Diffuse Lewy body disease; SCZ: Schizophrenia. (e–h) Exploration of demographic and clinical variables within subcohorts of samples from the three brain tissue sources, encompassing sex (e), genetically inferred ancestry (f), age (g), and disease status (h). NPS are not included in the disease count in (h). (i) Dendrogram of NPS based on their co-occurrence with three highlighted clusters.

Fig. 3

Analysis of genotyping data. (a) Counts and intersections among sources of genotyping data available for donors from the PsychAD cohort. (b) Distribution of genetic similarities estimated between combined genotype dataset and genotypes called from snRNA-seq data. (c) F-statistic from plink’s “check-sex” function plotted by reported sex (samples with known sex chromosome aneuploidies not shown). (d,e) The first two PCs of genetic ancestry were calculated separately for the PsychAD-MSSM genotype dataset of 882 samples (d) and for the combined dataset of 1,381 samples (e).

Fig. 4

Analysis of the snRNA-seq dataset. (a) Distribution of the number of nuclei across sample pools. Dashed line indicates the mean. (b) Distribution of nuclei to libraries within pools, ordered by nuclei count (top) and fraction of nuclei (bottom). Each replicate is depicted using two boxplots representing the nuclei distribution before (blue) and after QC (green). The center line (black) indicates the median, the box shows the interquartile range, and the whiskers indicate the highest/lowest values within 1.5 × the interquartile range. (c) Comparison of QC-passed nuclei counts between pairs of replicates from the same sequencing pools (Spearman’s ρ = 0.84). (d) Distribution of nuclei counts in samples that passed or failed QC (vertical line indicates the mean values). (e) UMAP visualization of snRNA-seq data. IN: inhibitory/GABAergic neurons, EN: excitatory/glutamatergic neurons, SMC: smooth muscle cells, VLMC: vascular leptomeningeal cells, PVM: perivascular macrophages, OPC: oligodendrocyte progenitor cells, Astro: astrocytes, Oligo: oligodendrocytes, Micro: Microglia, Endo: endothelial, Adaptive: adaptive immune cells, PC: Pericytes.

Fig. 5

Distribution of the age at death stratified by diagnosis. The diagnoses shown in this plot were intentionally selected to highlight age differences.