Fig. 7: scMORE distinguishes psychiatric disorder-associated eRegulons in a human cerebral organoid single-cell multiomic dataset.
a, Schematic representation of the scHOB database. scHOB curates and integrates data from over 1.5 million cells derived from various human organoids and their corresponding fetal tissues, including 83.4% scRNA-seq cells and 16.6% scATAC-seq cells. Example analysis was performed to demonstrate the utility of scMORE in identifying critical eRegulons associated with eight psychiatric disorders within organoid-based single-cell data. b, Distribution of regulon sizes identified by scMORE in human cerebral organoid single-cell multiomic data. Among these detected eRegulons, 71.66% consist of 20 or more genes. c, Heatmap illustrating scMORE-identified key eRegulons associated with eight psychiatric disorders across five brain cell types, categorized as three groups: glutamatergic neurons, GABAergic neurons and non-neuronal cells. **Denotes PCTS < 0.05, PGRS < 0.05, PTRS < 0.05 and TRS > 3. *Denotes PCTS < 0.05, PGRS < 0.05, PTRS < 0.05 and TRS > 1.5. Significance was assessed by an MC permutation test (one-sided, upper tail; n = 1,000 iterations; no multiple correction). d, Network visualization of disease-eRegulon associations within specific cellular contexts. This network-based analysis illustrates interactions between TF-eRegulons and psychiatric disorders in a cell type-specific manner. Diamonds represent psychiatric disorders (colored by cell type) and circles denote TF-eRegulons. e, Seven psychiatric disorder-specific eRegulons shared between glutamatergic and GABAergic neurons. f, Regulatory network of a representative eRegulon (FOXP2) that harbors strong genetic signals. Triangle denotes TF, diamonds represent psychiatric disorders, purple circles indicate target genes harboring GWAS loci, and light purple circles indicate risk SNPs. g, Functional enrichment analysis of the FOXP2-eRegulon based on GO BP using WebGestalt.
