Cell genomics最新文献

Single-cell RNA sequencing (scRNA-seq) provides an instantaneous snapshot of the transcriptional state of a cell, which results from the simultaneous activity of many cellular processes. In this issue of Cell Genomics, Chen et al.¹ describe the development of CellUntangler, a deep-learning-based model that allows the capture and filtering of multiple biological signals in scRNA-seq data.

Large-scale single-cell atlases have revealed many aging- and disease-associated cell types, yet these populations are often underrepresented in heterogeneous tissues, limiting detailed molecular analyses. To address this, we developed EnrichSci-a scalable, microfluidics-free platform that combines hybridization chain reaction RNA fluorescence in situ hybridization (FISH) with combinatorial indexing to profile single-nucleus transcriptomes of target cell types with full gene-body coverage. Applied to oligodendrocytes in the aging mouse brain, EnrichSci uncovered aging-associated molecular dynamics across distinct oligodendrocyte subtypes, revealing both shared and subtype-specific gene expression changes. Additionally, we identified aging-associated exon-level signatures missed by conventional gene-level analyses, highlighting post-transcriptional regulation as a critical dimension of cell-state dynamics in aging. By coupling transcript-guided enrichment with a scalable sequencing workflow, EnrichSci provides a versatile approach to decode dynamic regulatory landscapes in diverse cell types from complex tissues.

Sex differences in behaviors arise from variations in female and male nervous systems, yet the cellular and molecular bases of these differences remain poorly defined. Here, we employ an unbiased, single-cell transcriptomic approach to investigate how sex influences the adult Drosophila melanogaster brain. We demonstrate that sex differences do not result from large-scale transcriptional reprogramming, but rather from selective modifications within shared developmental lineages mediated by the sex-differentiating transcription factors Doublesex and Fruitless. We reveal, with unprecedented resolution, the extraordinary genetic diversity within these sexually dimorphic cell types and find that birth order represents a novel axis of sexual differentiation. Neuronal identity in the adult reflects spatiotemporal patterning and sex-specific survival, with female-biased neurons emerging early and male-biased neurons arising later. This pattern reframes dimorphic neurons as "paralogous" rather than "orthologous," suggesting sex leverages distinct developmental windows to build behavioral circuits, and highlights a role for exaptation in diversifying the brain.

Genome-wide association studies (GWASs) have identified over 50 lung cancer risk loci; however, the precise cellular context of these genetic mechanisms remains unclear due to limitations in bulk tissue expression quantitative trait locus (eQTL) analyses. Here, we present the largest single-cell eQTL (sc-eQTL) atlas of human lung tissue to date, profiling 222 donors using multiplexed single-cell RNA sequencing (scRNA-seq). We identified 4,341 independent eQTLs across 17 cell types, with over 60% of sc-eQTLs and 51% of eGenes being cell-type specific, and fewer than 52% were detectable in paired bulk datasets. Integration with GWASs for non-small cell lung cancer highlighted epithelial and immune cells as key contributors to genetic susceptibility, identifying 28 candidate genes within known risk loci and 24 in novel regions. Notably, 47% of established non-small cell lung cancer (NSCLC) susceptibility loci exhibited cell-type-specific pleiotropic genetic regulation. This study provides a valuable resource of lung sc-eQTLs and illuminates how genetic variation modulates gene expression in a cell-type-specific fashion, contributing to lung cancer susceptibility.

Histology images offer a cost-effective approach to predicting cellular phenotypes using spatial transcriptomics. However, existing methods struggle with individual gene expression accuracy and lack the capability to predict fine-grained transcriptional cell types. We present Hist2Cell, a vision graph-transformer framework to accurately resolve fine-grained cell types directly from histology images. Trained on human lung and breast cancer datasets, Hist2Cell predicts cell-type abundance with high accuracy (Pearson correlation over 0.80) and captures cellular colocalization. Moreover, it generalizes to large-scale The Cancer Genome Atlas (TCGA) cohorts without re-training, facilitating survival prediction by revealing distinct tissue microenvironments and cell type-patient mortality relationships. Thus, Hist2Cell enables cost-efficient analysis for large-scale spatial biology studies and precise cancer prognosis.

Understanding how a largely shared genome specifies distinct male and female behaviors is a central challenge in biology. Two recent papers show how sex-specific apoptosis interacts with neuron birth order in the Drosophila brain to sculpt male- and female-specific neural circuits from shared developmental templates.

Long-read whole-genome sequencing (LR-WGS) technologies enhance the discovery of structural variants (SVs) and tandem repeats (TRs). We performed LR-WGS on 267 individuals from 63 autism spectrum disorder (ASD) families and generated an integrated call set combining long- and short-read data. LR-WGS increased detection of gene-disrupting SVs and TRs by 33% and 38%, respectively, and enabled identification of novel exonic de novo germline and somatic SVs. We observed complex SV patterns, including a class of nested duplication-deletion events. By joint analysis of phased genetic variation and DNA methylation, we identified deletions of imprinted genes and demonstrated the effect of intermediate TR expansions (35-54 CGG) on the methylation of FMR1 promoter. Rare SVs, TRs, and damaging SNVs together accounted for 7.4% (95% confidence interval [CI], 2.7%-17%) of the heritability of ASD. These findings demonstrate how LR-WGS can resolve complex genetic variation and its functional consequences and regulatory effects in a single assay.

CRISPR-based genome editing has revolutionized functional genomics, enabling thousands of perturbations to be concurrently assayed in single experiments. However, for methods such as saturation genome editing (SGE), which aims to generate and assay libraries of point mutations, a challenge is that only one region (e.g., one exon) is studied per experiment. Here, we describe prime-SGE, a prime editing-based framework in which libraries of specific point mutations are installed into genes throughout the genome and then functionally assessed by sequencing of prime editing guide RNAs (pegRNAs) rather than the mutations themselves. We apply prime-SGE in two cell lines to assay thousands of point mutations in eight oncogenes for their ability to confer drug resistance to four tyrosine kinase inhibitors. Our prime-SGE strategy, combined with ongoing improvements in prime editing efficiency, opens the door to efficient positive selection screens of large numbers of point mutations at locations throughout the genome.

Standard techniques for detecting genomic rearrangements in formalin-fixed paraffin-embedded (FFPE) biopsies have important limitations. We performed FFPE-compatible Hi-C on 44 clinical biopsies comprising large B cell lymphomas (n = 18), plasma cell neoplasms (n = 14), and other diverse lymphoid cancers, identifying consistent topological differences between malignant B cell and plasma cell states. Hi-C detected expected oncogene rearrangements at high concordance with fluorescence in situ hybridization (FISH) and supported enhancer hijacking in recurrent rearrangements of BCL2, CCND1, and MYC plus unanticipated variants involving homologous loci. Hi-C identified unanticipated non-coding rearrangements involving PD-1 ligand genes and other loci of potential therapeutic relevance, distinguished between functionally divergent classes of BCL6 rearrangements, and provided topological information supporting interpretation of variant MYC rearrangements. Hi-C revealed disease-selective MYC locus topological features that correlated with disease-selective MYC locus enhancers and rearrangement breakpoint distributions. FFPE-compatible Hi-C detects oncogene rearrangements and their topological consequences at genome-wide scale, finding clinically relevant drivers missed by standard approaches.

Human height is a model for the genetic analysis of complex traits, and recent studies suggest the presence of thousands of common genetic variant associations and hundreds of low-frequency/rare variants. Here, we develop a new algorithmic paradigm based on approximate message passing (genomic vector approximate message passing [gVAMP]) for identifying DNA sequence variants associated with complex traits and common diseases in large-scale whole-genome sequencing (WGS) data. We show that gVAMP accurately localizes associations to variants with the correct frequency and position in the DNA, outperforming existing fine-mapping methods in selecting the appropriate genetic variants within WGS data. We then apply gVAMP to jointly model the relationship of tens of millions of WGS variants with human height in hundreds of thousands of UK Biobank individuals. We identify 59 rare variants and gene burden scores alongside many hundreds of DNA regions containing common variant associations and show that understanding the genetic basis of complex traits will require the joint analysis of hundreds of millions of variables measured on millions of people. The polygenic risk scores obtained from gVAMP have high accuracy (including a prediction accuracy of ∼46% for human height) and outperform current methods for downstream tasks such as mixed linear model association testing across 13 UK Biobank traits. In conclusion, gVAMP offers a scalable foundation for a wider range of analyses in WGS data.