Data integration of bulk and single-cell transcriptomics from cerebral organoids and post-mortem brains to identify cell types and cell type specific driver genes in autism

Human-derived cerebral organoids demonstrate great promise for identifying cell types and cell type specific molecular processes perturbed by genetic variants associated with neuropsychiatric and neurodevelopmental disorders, which are notoriously challenging to study using animal models. However, considerable challenges remain in achieving robust, scalable and generalizable phenotyping of organoids to discover cell types and cell type specific genes. We perform RNA sequencing on 71 samples comprising 1,420 cerebral organoids from 25 donors, and describe a framework (Orgo-Seq) to integrate bulk RNA and single-cell RNA sequence data from human post-mortem brains and cerebral organoids, for the identification of cell types and cell type specific individual genes. We apply Orgo-Seq for two autism-associated loci: 16p11.2 deletions and 15q11-13 duplications, and identify neuroepithelial cells as critical cell types for 16p11.2 deletions, and discover novel and previously reported cell type specific driver genes. Finally, we validated our results that mutations in the KCTD13 gene in the 16p11.2 locus lead to imbalances in the proportion of neuroepithelial cells, using CRISPR/Cas9-edited mosaic organoids. Our work presents a quantitative technological framework to integrate multiple transcriptomics datasets to identify cell types and cell type specific driver genes associated with complex diseases using cerebral organoids.