New York Genome Center

Whole genome sequencing (WGS) studies play a pivotal role in studying the genetic underpinnings of human diseases and traits. High quality and reproducible variant calling is the cornerstone for the success of downstream analyses, including WGS association studies and polygenic risk prediction. This paper compares the data quality, performance, and concordance of two widely used WGS variant callers, the Genome Analysis Toolkit (GATK) and Variant Tool set that discovers short variants (VT), using 60532 multi-ancestry whole genomes sequenced by the Centers for Common Disease Genomics (CCDGs) of the NHGRI Genome Sequencing Program. Our findings show that both QCed GATK and VT pipelines yield highly consistent and reliable called Single Nucleotide Variants (SNVs) in large-scale WGS studies, supporting their agreements in joint variants calling. However, the two pipelines exhibit greater discrepancies in calling insertions and deletions (INDELs).

Suicide is a significant public health problem that usually co-occurs with major psychiatric disorders. Suicidal behaviors have heritability of 30-50%, and the largest genome-wide association studies identified 12 loci linked to suicide attempts (SA). These findings indicate shared genetic architecture among SA and psychiatric disorders. We analyzed public GWAS summary statistics of SA, major depressive disorder (MDD), bipolar disorder (BIP), schizophrenia (SCZ), and attention deficit hyperactivity disorder (ADHD) to quantify genetic overlap using statistical genetics methods: MiXeR for polygenic overlap, LAVA for locus-specific genetic correlations, and HyPrColoc for multi-trait colocalization. MTAG and conditional false discovery rate (condFDR) identified SA- associated loci, while conjunctional false discovery rate (conjFDR) identified shared loci with psychiatric disorders. Additionally, we used polygenic risk scores (PRS) calculated in the UK Biobank to validate associations between genetic liabilities of psychiatric disorders and SA. SA involve approximately 6.9k (SD = 1.5k) risk variants with substantial overlap between psychiatric disorders, ranging from 53.2% with SCZ to 82.1% with BIP. Using MTAG, condFDR, and conjFDR, we identified 14 and 48 novel risk loci for SA, respectively. Through conjFDR, we identified 78 loci associated with SA and major psychiatric disorders, including one locus shared across all five traits. Genes linked to SA were enriched in synaptic components and signaling pathways. Despite significant genetic overlap, the SA PRS was the single strongest predictor of SA, followed by the MDD and ADHD PRS. The genetic overlap reflects potential common comorbidities complicating the identification of biological processes unique to SA, instead reflecting a complex genetic framework shared with psychiatric disorders.

Calmodulin is an intracellular Ca2+ sensor that regulates numerous cellular processes through binding effector proteins and changing their activity. Humans have three calmodulin paralogs, CALM1, CALM2 and CALM3, encoding identical proteins, and missense variants in all three are included in the ACMG recommendations for reporting of secondary findings related to early onset sudden cardiac death caused by arrhythmias long QT syndrome and CPVT. Recently, a subset of individuals with de novo variants in calmodulin genes were described who also presented with neurologic phenotypes. Here we report two individuals with the same de novo variant c.419A > T in CALM1 or CALM2 that may generate a 5' splice donor gain and/or a missense change p.E140V. These individuals share hypotonia, motor delay, intellectual disability, and abnormal electroencephalograms but lack cardiac arrhythmia/electrocardiogram abnormalities of previously described CALM-associated disease, suggesting the variant causes phenotypic expansion beyond the known Mendelian phenotypes. RNA-seq of the CALM1 proband blood's sample revealed that most transcripts from the variant allele showed usage of the new splice site or intron retention, without NMD, resulting in frameshifted C-terminal truncations, while a minority resulted in production of the p.E140V missense protein. We modeled the CALM1/2 p.E140V variant as well as a known arrhythmia variant, CALM1 p.E141G, using the C. elegans ortholog cmd-1. We found that cmd-1 E140V displayed both qualitative and quantitative differences in phenotype from E141G, indicating distinct genetic mechanisms. Together, these findings support CALM1/2 phenotypic expansion, with c.419A > T p.E140V resulting in neurologic, but not arrhythmia phenotypes.

Transcript diversity including splicing and alternative 3' end usage is crucial for cellular identity and adaptation, yet its spatial coordination remains poorly understood. Here we present SPLISOSM (spatial isoform statistical modeling), a method for detecting isoform-resolution patterns from spatial transcriptomics data. SPLISOSM uses multivariate testing with nonparametric kernels to account for spot-level and isoform-level dependencies, achieving high statistical power on sparse data. In the mouse brain, we identify over 1,000 spatially variable transcript diversity events, primarily in synaptic signaling pathways linked to neuropsychiatric disorders, and uncover both known and previously unknown regulatory relationships with region-specific RNA binding proteins. We further show that these patterns are evolutionarily conserved between mouse and human prefrontal cortex. Analysis of human glioblastoma highlights pervasive transcript diversity in antigen presentation and adhesion genes associated with specific microenvironmental conditions. Together, we present a comprehensive spatial splicing analysis in the brain under normal and neoplastic conditions.

Therapies that harness the immune system to target and eliminate tumour cells have revolutionized cancer care. Immune checkpoint blockade (ICB), which boosts the anti-tumour immune response by inhibiting negative regulators of T cell activation1–3, is remarkably successful in a subset of cancer patients. Yet a significant proportion do not respond to treatment, emphasizing the need to understand factors influencing the therapeutic efficacy of ICB4–9. The gut microbiota, consisting of trillions of microorganisms residing in the gastrointestinal tract, has emerged as a critical determinant of immune function and response to cancer immunotherapy, with several studies demonstrating association of microbiota composition with clinical response10–16. However, a mechanistic understanding of how gut commensal bacteria influence the efficacy of ICB remains elusive. Here we use a gut commensal microorganism, segmented filamentous bacteria (SFB), which induces an antigen-specific T helper 17 (TH17) cell effector program in the small intestine lamina propria (SILP)17, to investigate how colonization with this microbe affects the efficacy of ICB in restraining distal growth of tumours sharing antigen with SFB. We find that anti-programmed cell death protein 1 (PD-1) treatment effectively inhibits the growth of implanted SFB antigen-expressing melanoma only if mice are colonized with SFB. Through T cell receptor (TCR) clonal lineage tracing, fate mapping and peptide–major histocompatability complex (MHC) tetramer staining, we identify tumour-associated SFB-specific T helper 1 (TH1)-like cells derived from the homeostatic TH17 cells induced by SFB colonization in the SILP. These gut-educated ex-TH17 cells produce high levels of the pro-inflammatory cytokines interferon (IFN)-γ and tumour necrosis factor (TNF) within the tumour microenvironment (TME), enhancing antigen presentation and promoting recruitment, expansion and effector functions of CD8+ tumour-infiltrating cytotoxic lymphocytes and thereby enabling anti-PD-1-mediated tumour control. Conditional ablation of SFB-induced IL-17A+CD4+ T cells, precursors of tumour-associated TH1-like cells, abolishes anti-PD-1-mediated tumour control and markedly impairs tumour-specific CD8+ T cell recruitment and effector function within the TME. Our data, as a proof of principle, define a cellular pathway by which a single, defined intestinal commensal imprints T cell plasticity that potentiates PD-1 blockade, and indicate targeted modulation of the microbiota as a strategy to broaden ICB efficacy.

Translating human genomic discoveries into mechanistic insights requires linking genetic variations to candidate genes and their causal functional phenotypes. Genome-wide association studies have consistently identified LIPA (lipase A, lysosomal acid type) as a risk locus for coronary artery disease, with previous analyses prioritising LIPA as a likely causal gene. However, functional studies elucidating causal variants, regulatory mechanisms, target cell types, and their causal impact on atherosclerosis have been lacking. This study aims to address this gap by establishing the variant-to-function relationship at the LIPA locus. Post-genome-wide association study pipelines and molecular biology techniques, including expression quantitative trait loci analysis, Tri-HiC, luciferase assay, CRISPRi, allele-specific binding, motif analysis, and electrophoretic mobility shift assay, were used to link functional variants to target genes and define the direction of their regulatory effects in causal cell types. To determine how increased myeloid LIPA impacts atherosclerosis, myeloid-specific Lipa overexpression mice on an Ldlr-/- background were generated. Coronary artery disease-risk alleles in the LIPA locus increase LIPA expression and enzyme activity specifically in monocytes/macrophages by enhancing PU.1 binding to an intronic enhancer region that interacts with the LIPA promoter. Myeloid-specific Lipa overexpression in Ldlr-/- mice fed a western diet resulted in larger atherosclerotic lesions, accompanied by altered macrophage function, characterized by increased accumulation of lesional macrophages derived from circulating monocytes, reduced neutral lipid content, and up-regulation of integrin and extracellular matrix pathway genes. The work establishes a direct causal link between LIPA-risk alleles and increased monocyte/macrophage LIPA that exacerbates atherosclerosis, bridging human functional genomic evidence to the mechanistic understanding of coronary artery disease.

Extrachromosomal DNAs (ecDNAs) are circular DNA molecules prevalent in human cancers that drive tumor evolution and drug resistance. Their circular topology, which disrupts topological domains and rewires regulatory circuits, has typically been studied via pairwise interactions. Here we develop ec3D, a computational method for reconstructing three-dimensional ecDNA structures from Hi-C data. Given a candidate ecDNA sequence and whole-genome Hi-C data, ec3D reconstructs spatial structures by maximizing the Poisson likelihood of observed interactions. We validate ec3D using simulated structures, previously characterized cancer cell lines, and microscopy imaging. Our reconstructions reveal that ecDNAs occupy spherical configurations and mediate unique long-range regulatory interactions involved in gene regulation. Through algorithmic innovations, ec3D can resolve complex structures with duplicated segments, identify multi-way interactions, and identify potential intermolecular (trans) interactions. Our findings provide insights into how ecDNA’s spatial organization bypasses normal chromosomal constraints and contributes to increased oncogene expression.

BackgroundTranscriptomic analysis is common in large cohort studies but is generally restricted to cells in blood, which limits inferences about organs of interest, and direct organ sampling is mostly infeasible in large cohorts. New techniques for RNA-seq from noninvasive biosamples may provide the opportunity to profile transcriptomes of additional tissues for more organ-relevant insights at scale. We investigated the feasibility and utility of hair follicle gene expression profiling in a multi-center study of chronic obstructive pulmonary disease (COPD).MethodsBulk RNA-seq was performed on hair follicles collected in the SubPopulations and InteRmediate Outcome Measures in COPD Study (SPIROMICS), a multi-center longitudinal study of COPD (n = 97). The resulting hair follicle gene expression data were characterized and compared both to gene expression in whole blood and bronchial epithelium previously measured in SPIROMICS and to Genotype-Tissue Expression (GTEx) project tissue gene expression by principal component analysis and single-sample gene enrichment analysis, used to estimate hair follicle cell type proportions, and tested for association with disease-relevant lung phenotypes. eQTL discovery was also performed and colocalization with a genome-wide association study for lung function was tested.ResultsHair follicles reliably produced transcriptomic data of sufficient quality and number for cell type composition, which revealed mostly epithelial and fibroblast cells. Comparison to other tissues previously profiled in SPIROMICS and GTEx project demonstrated transcriptomes from hair follicles were much more similar to those from lung parenchyma than blood. Combining these data with rich clinical, imaging, and genomic profiling in SPIROMICS, we found that they provided an attractive approach for discovery of associations with complex lung phenotypes, particularly of the airways. Finally, we investigated hair follicle genetic architecture through expression quantitative trait locus (eQTL) discovery and demonstrated better colocalization with lung-related genetic associations than blood.ConclusionHere, we demonstrated that RNA-seq applied to hair follicle transcriptomic profiling can be scaled up successfully in a multi-center study to yield inferences not available from blood transcriptomics.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12920-025-02270-w.