Single-cell genomics and regulatory networks for 388 human brains.

Single-cell genomics is a powerful tool for studying heterogeneous tissues such as the brain. Yet little is understood about how genetic variants influence cell-level gene expression. Addressing this, we uniformly processed single-nuclei, multiomics datasets into a resource comprising >2.8 million nuclei from the prefrontal cortex across 388 individuals. For 28 cell types, we assessed population-level variation in expression and chromatin across gene families and drug targets. We identified >550,000 cell type-specific regulatory elements and >1.4 million single-cell expression quantitative trait loci, which we used to build cell-type regulatory and cell-to-cell communication networks. These networks manifest cellular changes in aging and neuropsychiatric disorders. We further constructed an integrative model accurately imputing single-cell expression and simulating perturbations; the model prioritized ~250 disease-risk genes and drug targets with associated cell types.

Single-cell multi-cohort dissection of the schizophrenia transcriptome.

The complexity and heterogeneity of schizophrenia have hindered mechanistic elucidation and the development of more effective therapies. Here, we performed single-cell dissection of schizophrenia-associated transcriptomic changes in the human prefrontal cortex across 140 individuals in two independent cohorts. Excitatory neurons were the most affected cell group, with transcriptional changes converging on neurodevelopment and synapse-related molecular pathways. Transcriptional alterations included known genetic risk factors, suggesting convergence of rare and common genomic variants on neuronal population-specific alterations in schizophrenia. Based on the magnitude of schizophrenia-associated transcriptional change, we identified two populations of individuals with schizophrenia marked by expression of specific excitatory and inhibitory neuronal cell states. This single-cell atlas links transcriptomic changes to etiological genetic risk factors, contextualizing established knowledge within the human cortical cytoarchitecture and facilitating mechanistic understanding of schizophrenia pathophysiology and heterogeneity.

Genetic regulation of cell type-specific chromatin accessibility shapes brain disease etiology.

Nucleotide variants in cell type-specific gene regulatory elements in the human brain are risk factors for human disease. We measured chromatin accessibility in 1932 aliquots of sorted neurons and non-neurons from 616 human postmortem brains and identified 34,539 open chromatin regions with chromatin accessibility quantitative trait loci (caQTLs). Only 10.4% of caQTLs are shared between neurons and non-neurons, which supports cell type-specific genetic regulation of the brain regulome. Incorporating allele-specific chromatin accessibility improves statistical fine-mapping and refines molecular mechanisms that underlie disease risk. Using massively parallel reporter assays in induced excitatory neurons, we screened 19,893 brain QTLs and identified the functional impact of 476 regulatory variants. Combined, this comprehensive resource captures variation in the human brain regulome and provides insights into disease etiology.

Evaluating performance and applications of sample-wise cell deconvolution methods on human brain transcriptomic data.

Sample-wise deconvolution methods estimate cell-type proportions and gene expressions in bulk tissue samples, yet their performance and biological applications remain unexplored, particularly in human brain transcriptomic data. Here, nine deconvolution methods were evaluated with sample-matched data from bulk tissue RNA sequencing (RNA-seq), single-cell/nuclei (sc/sn) RNA-seq, and immunohistochemistry. A total of 1,130,767 nuclei per cells from 149 adult postmortem brains and 72 organoid samples were used. The results showed the best performance of dtangle for estimating cell proportions and bMIND for estimating sample-wise cell-type gene expressions. For eight brain cell types, 25,273 cell-type eQTLs were identified with deconvoluted expressions (decon-eQTLs). The results showed that decon-eQTLs explained more schizophrenia GWAS heritability than bulk tissue or single-cell eQTLs did alone. Differential gene expressions associated with Alzheimer's disease, schizophrenia, and brain development were also examined using the deconvoluted data. Our findings, which were replicated in bulk tissue and single-cell data, provided insights into the biological applications of deconvoluted data in multiple brain disorders.

Single-cell genomics and regulatory networks for 388 human brains.

Single-cell genomics is a powerful tool for studying heterogeneous tissues such as the brain. Yet, little is understood about how genetic variants influence cell-level gene expression. Addressing this, we uniformly processed single-nuclei, multi-omics datasets into a resource comprising >2.8M nuclei from the prefrontal cortex across 388 individuals. For 28 cell types, we assessed population-level variation in expression and chromatin across gene families and drug targets. We identified >550K cell-type-specific regulatory elements and >1.4M single-cell expression-quantitative-trait loci, which we used to build cell-type regulatory and cell-to-cell communication networks. These networks manifest cellular changes in aging and neuropsychiatric disorders. We further constructed an integrative model accurately imputing single-cell expression and simulating perturbations; the model prioritized ~250 disease-risk genes and drug targets with associated cell types.

Sex differences in brain cell-type specific chromatin accessibility in schizophrenia.

Our understanding of the sex-specific role of the non-coding genome in serious mental illness remains largely incomplete. To address this gap, we explored sex differences in 1,393 chromatin accessibility profiles, derived from neuronal and non-neuronal nuclei of two distinct cortical regions from 234 cases with serious mental illness and 235 controls. We identified sex-specific enhancer-promoter interactions and showed that they regulate genes involved in X-chromosome inactivation (XCI). Examining chromosomal conformation allowed us to identify sex-specific cis- and trans-regulatory domains (CRDs and TRDs). Co-localization of sex-specific TRDs with schizophrenia common risk variants pinpointed male-specific regulatory regions controlling a number of metabolic pathways. Additionally, enhancers from female-specific TRDs were found to regulate two genes known to escape XCI, (XIST and JPX), underlying the importance of TRDs in deciphering sex differences in schizophrenia. Overall, these findings provide extensive characterization of sex differences in the brain epigenome and disease-associated regulomes.

Alzheimer's disease transcriptional landscape in ex-vivo human microglia.

Microglia are resident immune cells of the brain and are implicated in the etiology of Alzheimer's Disease (AD) and other diseases. Yet the cellular and molecular processes regulating their function throughout the course of the disease are poorly understood. Here, we present the transcriptional landscape of primary microglia from 189 human postmortem brains, including 58 healthy aging individuals and 131 with a range of disease phenotypes, including 63 patients representing the full spectrum of clinical and pathological severity of AD. We identified transcriptional changes associated with multiple AD phenotypes, capturing the severity of dementia and neuropathological lesions. Transcript-level analyses identified additional genes with heterogeneous isoform usage and AD phenotypes. We identified changes in gene-gene coordination in AD, dysregulation of co-expression modules, and disease subtypes with distinct gene expression. Taken together, these data further our understanding of the key role of microglia in AD biology and nominate candidates for therapeutic intervention.

Convergence of the dysregulated regulome in schizophrenia with polygenic risk and evolutionarily constrained enhancers.

Enhancers play an essential role in the etiology of schizophrenia; however, the dysregulation of enhancer activity and its impact on the regulome in schizophrenia remains understudied. To address this gap in our knowledge, we assessed enhancer and gene expression in 1,382 brain samples comprising cases with schizophrenia and unaffected controls. Dysregulation of enhancer expression was concordant with changes in gene expression, and was more closely associated with schizophrenia polygenic risk, suggesting that enhancer dysregulation is proximal to the genetic etiology of the disease. Modeling the shared variance of cis-coordinated genes and enhancers revealed a gene regulatory program that was highly associated with genetic vulnerability to schizophrenia. By integrating coordinated factors with evolutionary constraints, we found that enhancers acquired during human evolution are more likely to regulate genes that are implicated in neuropsychiatric disorders and, thus, hold potential as therapeutic targets. Our analysis provides a systematic view of regulome dysregulation in schizophrenia and highlights its convergence with schizophrenia polygenic risk and human-gained enhancers.

Activity-Dependent Transcriptional Program in NGN2+ Neurons Enriched for Genetic Risk for Brain-Related Disorders.

BACKGROUND: Converging evidence from large-scale genetic and postmortem studies highlights the role of aberrant neurotransmission and genetic regulation in brain-related disorders. However, identifying neuronal activity-regulated transcriptional programs in the human brain and understanding how changes contribute to disease remain challenging. METHODS: To better understand how the activity-dependent regulome contributes to risk for brain-related disorders, we profiled the transcriptomic and epigenomic changes following neuronal depolarization in human induced pluripotent stem cell-derived glutamatergic neurons (NGN2) from 6 patients with schizophrenia and 5 control participants. RESULTS: Multiomic data integration associated global patterns of chromatin accessibility with gene expression and identified enhancer-promoter interactions in glutamatergic neurons. Within 1 hour of potassium chloride-induced depolarization, independent of diagnosis, glutamatergic neurons displayed substantial activity-dependent changes in the expression of genes regulating synaptic function. Depolarization-induced changes in the regulome revealed significant heritability enrichment for schizophrenia and Parkinson's disease, adding to mounting evidence that sequence variation within activation-dependent regulatory elements contributes to the genetic risk for brain-related disorders. Gene coexpression network analysis elucidated interactions among activity-dependent and disease-associated genes and pointed to a key driver (NAV3) that interacted with multiple genes involved in axon guidance. CONCLUSIONS: Overall, we demonstrated that deciphering the activity-dependent regulome in glutamatergic neurons reveals novel targets for advanced diagnosis and therapy.

Long-read RNA-seq atlas of novel microglia isoforms elucidates disease-associated genetic regulation of splicing.

Microglia, the innate immune cells of the central nervous system, have been genetically implicated in multiple neurodegenerative diseases. We previously mapped the genetic regulation of gene expression and mRNA splicing in human microglia, identifying several loci where common genetic variants in microglia-specific regulatory elements explain disease risk loci identified by GWAS. However, identifying genetic effects on splicing has been challenging due to the use of short sequencing reads to identify causal isoforms. Here we present the isoform-centric microglia genomic atlas (isoMiGA) which leverages the power of long-read RNA-seq to identify 35,879 novel microglia isoforms. We show that the novel microglia isoforms are involved in stimulation response and brain region specificity. We then quantified the expression of both known and novel isoforms in a multi-ethnic meta-analysis of 555 human microglia short-read RNA-seq samples from 391 donors, the largest to date, and found associations with genetic risk loci in Alzheimer's disease and Parkinson's disease. We nominate several loci that may act through complex changes in isoform and splice site usage.

The neuronal chromatin landscape in adult schizophrenia brains is linked to early fetal development.

Non-coding variants increase risk of neuropsychiatric disease. However, our understanding of the cell-type specific role of the non-coding genome in disease is incomplete. We performed population scale (N=1,393) chromatin accessibility profiling of neurons and non-neurons from two neocortical brain regions: the anterior cingulate cortex and dorsolateral prefrontal cortex. Across both regions, we observed notable differences in neuronal chromatin accessibility between schizophrenia cases and controls. A per-sample disease pseudotime was positively associated with genetic liability for schizophrenia. Organizing chromatin into cis- and trans-regulatory domains, identified a prominent neuronal trans-regulatory domain (TRD1) active in immature glutamatergic neurons during fetal development. Polygenic risk score analysis using genetic variants within chromatin accessibility of TRD1 successfully predicted susceptibility to schizophrenia in the Million Veteran Program cohort. Overall, we present the most extensive resource to date of chromatin accessibility in the human cortex, yielding insights into the cell-type specific etiology of schizophrenia.

DeepGAMI: deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype-phenotype prediction.

BACKGROUND: Genotypes are strongly associated with disease phenotypes, particularly in brain disorders. However, the molecular and cellular mechanisms behind this association remain elusive. With emerging multimodal data for these mechanisms, machine learning methods can be applied for phenotype prediction at different scales, but due to the black-box nature of machine learning, integrating these modalities and interpreting biological mechanisms can be challenging. Additionally, the partial availability of these multimodal data presents a challenge in developing these predictive models. METHOD: To address these challenges, we developed DeepGAMI, an interpretable neural network model to improve genotype-phenotype prediction from multimodal data. DeepGAMI leverages functional genomic information, such as eQTLs and gene regulation, to guide neural network connections. Additionally, it includes an auxiliary learning layer for cross-modal imputation allowing the imputation of latent features of missing modalities and thus predicting phenotypes from a single modality. Finally, DeepGAMI uses integrated gradient to prioritize multimodal features for various phenotypes. RESULTS: We applied DeepGAMI to several multimodal datasets including genotype and bulk and cell-type gene expression data in brain diseases, and gene expression and electrophysiology data of mouse neuronal cells. Using cross-validation and independent validation, DeepGAMI outperformed existing methods for classifying disease types, and cellular and clinical phenotypes, even using single modalities (e.g., AUC score of 0.79 for Schizophrenia and 0.73 for cognitive impairment in Alzheimer's disease). CONCLUSION: We demonstrated that DeepGAMI improves phenotype prediction and prioritizes phenotypic features and networks in multiple multimodal datasets in complex brains and brain diseases. Also, it prioritized disease-associated variants, genes, and regulatory networks linked to different phenotypes, providing novel insights into the interpretation of gene regulatory mechanisms. DeepGAMI is open-source and available for general use.

The neuronal chromatin landscape in adult schizophrenia brains is linked to early fetal development.

Non-coding variants increase risk of neuropsychiatric disease. However, our understanding of the cell-type specific role of the non-coding genome in disease is incomplete. We performed population scale (N=1,393) chromatin accessibility profiling of neurons and non-neurons from two neocortical brain regions: the anterior cingulate cortex and dorsolateral prefrontal cortex. Across both regions, we observed notable differences in neuronal chromatin accessibility between schizophrenia cases and controls. A per-sample disease pseudotime was positively associated with genetic liability for schizophrenia. Organizing chromatin into cis- and trans-regulatory domains, identified a prominent neuronal trans-regulatory domain (TRD1) active in immature glutamatergic neurons during fetal development. Polygenic risk score analysis using genetic variants within chromatin accessibility of TRD1 successfully predicted susceptibility to schizophrenia in the Million Veteran Program cohort. Overall, we present the most extensive resource to date of chromatin accessibility in the human cortex, yielding insights into the cell-type specific etiology of schizophrenia.

Genome-wide analysis of a model-derived binge eating disorder phenotype identifies risk loci and implicates iron metabolism.

Binge eating disorder (BED) is the most common eating disorder, yet its genetic architecture remains largely unknown. Studying BED is challenging because it is often comorbid with obesity, a common and highly polygenic trait, and it is underdiagnosed in biobank data sets. To address this limitation, we apply a supervised machine-learning approach (using 822 cases of individuals diagnosed with BED) to estimate the probability of each individual having BED based on electronic medical records from the Million Veteran Program. We perform a genome-wide association study of individuals of African (n = 77,574) and European (n = 285,138) ancestry while controlling for body mass index to identify three independent loci near the HFE, MCHR2 and LRP11 genes and suggest APOE as a risk gene for BED. We identify shared heritability between BED and several neuropsychiatric traits, and implicate iron metabolism in the pathophysiology of BED. Overall, our findings provide insights into the genetics underlying BED and suggest directions for future translational research.

Efficient differential expression analysis of large-scale single cell transcriptomics data using dreamlet.

Advances in single-cell and -nucleus transcriptomics have enabled generation of increasingly large-scale datasets from hundreds of subjects and millions of cells. These studies promise to give unprecedented insight into the cell type specific biology of human disease. Yet performing differential expression analyses across subjects remains difficult due to challenges in statistical modeling of these complex studies and scaling analyses to large datasets. Our open-source R package dreamlet (DiseaseNeurogenomics.github.io/dreamlet) uses a pseudobulk approach based on precision-weighted linear mixed models to identify genes differentially expressed with traits across subjects for each cell cluster. Designed for data from large cohorts, dreamlet is substantially faster and uses less memory than existing workflows, while supporting complex statistical models and controlling the false positive rate. We demonstrate computational and statistical performance on published datasets, and a novel dataset of 1.4M single nuclei from postmortem brains of 150 Alzheimer's disease cases and 149 controls.

Genetic regulation of cell-type specific chromatin accessibility shapes the etiology of brain diseases.

Nucleotide variants in cell type-specific gene regulatory elements in the human brain are major risk factors of human disease. We measured chromatin accessibility in sorted neurons and glia from 1,932 samples of human postmortem brain and identified 34,539 open chromatin regions with chromatin accessibility quantitative trait loci (caQTL). Only 10.4% of caQTL are shared between neurons and glia, supporting the cell type specificity of genetic regulation of the brain regulome. Incorporating allele specific chromatin accessibility improves statistical fine-mapping and refines molecular mechanisms underlying disease risk. Using massively parallel reporter assays in induced excitatory neurons, we screened 19,893 brain QTLs, identifying the functional impact of 476 regulatory variants. Combined, this comprehensive resource captures variation in the human brain regulome and provides novel insights into brain disease etiology.

Efficient differential expression analysis of large-scale single cell transcriptomics data using dreamlet.

Advances in single-cell and -nucleus transcriptomics have enabled generation of increasingly large-scale datasets from hundreds of subjects and millions of cells. These studies promise to give unprecedented insight into the cell type specific biology of human disease. Yet performing differential expression analyses across subjects remains difficult due to challenges in statistical modeling of these complex studies and scaling analyses to large datasets. Our open-source R package dreamlet (DiseaseNeurogenomics.github.io/dreamlet) uses a pseudobulk approach based on precision-weighted linear mixed models to identify genes differentially expressed with traits across subjects for each cell cluster. Designed for data from large cohorts, dreamlet is substantially faster and uses less memory than existing workflows, while supporting complex statistical models and controlling the false positive rate. We demonstrate computational and statistical performance on published datasets, and a novel dataset of 1.4M single nuclei from postmortem brains of 150 Alzheimer's disease cases and 149 controls.

Evaluating performance and applications of sample-wise cell deconvolution methods on human brain transcriptomic data.

Sample-wise deconvolution methods have been developed to estimate cell-type proportions and gene expressions in bulk-tissue samples. However, the performance of these methods and their biological applications has not been evaluated, particularly on human brain transcriptomic data. Here, nine deconvolution methods were evaluated with sample-matched data from bulk-tissue RNAseq, single-cell/nuclei (sc/sn) RNAseq, and immunohistochemistry. A total of 1,130,767 nuclei/cells from 149 adult postmortem brains and 72 organoid samples were used. The results showed the best performance of dtangle for estimating cell proportions and bMIND for estimating sample-wise cell-type gene expression. For eight brain cell types, 25,273 cell-type eQTLs were identified with deconvoluted expressions (decon-eQTLs). The results showed that decon-eQTLs explained more schizophrenia GWAS heritability than bulk-tissue or single-cell eQTLs alone. Differential gene expression associated with multiple phenotypes were also examined using the deconvoluted data. Our findings, which were replicated in bulk-tissue RNAseq and sc/snRNAseq data, provided new insights into the biological applications of deconvoluted data.

Detecting and Adjusting for Hidden Biases due to Phenotype Misclassification in Genome-Wide Association Studies.

With the advent of healthcare-based genotyped biobanks, genome-wide association studies (GWAS) leverage larger sample sizes, incorporate patients with diverse ancestries and introduce noisier phenotypic definitions. Yet the extent and impact of phenotypic misclassification on large-scale datasets is not currently well understood due to a lack of statistical methods to estimate relevant parameters from empirical data. Here, we develop a statistical method and scalable software, PheMED, Phenotypic Measurement of Effective Dilution, to quantify phenotypic misclassification across GWAS using only summary statistics. We illustrate how the parameters estimated by PheMED relate to the negative and positive predictive value of the labeled phenotype, compared to ground truth, and how misclassification of the phenotype yields diluted effect-sizes of variant-phenotype associations. Furthermore, we apply our methodology to detect multiple instances of statistically significant dilution in real-world data. We demonstrate how effective dilution biases downstream GWAS replication and heritability analyses despite utilizing current best practices, and provide a dilution-aware meta-analysis approach that outperforms existing methods. Consequently, we anticipate that PheMED will be a valuable tool for researchers to address phenotypic data quality issues both within and across cohorts.