The rich ethnolinguistic and sociocultural differences that exist in India offers a unique opportunity to study human diversity. With the whole genomes of 10,000 healthy and unrelated Indians from 83 populations, the GenomeIndia project captures the genetic diversity of one of the highly underrepresented populations in the global genomics landscape.

The biological significance of somatic mutations in rare genetic liver diseases has remained elusive. A study using genomic sequencing paired with experimental cell-based assays has shed light on the selective advantage and accumulation of somatic variants in SERPINA1 found in liver explants from patients with alpha-1 anti-trypsin deficiency.

Borzoi is a deep learning model that predicts RNA sequencing coverage across each exon of every human gene, across different cells and tissues, based on DNA sequence alone.

Modern genome-wide association study (GWAS) datasets are biobank scale and are distributed across isolated repositories, owing to privacy concerns and regulations. Secure federated GWAS leverages cryptographic tools to enable joint analysis of these siloed datasets, thereby enhancing a range of common analyses while providing formal guarantees of data privacy.

We identify that signal-dependent and ligand-dependent enhancer activation programs require release of a shared enhancer RNA (eRNA) transcription checkpoint. Further mechanistic evidence reveals signal-dependent activation of P-TEFb on enhancers, which licenses the enhancer transcriptional program by overcoming pausing of transcription and promoting eRNA elongation.

Japanese soybeans have been bred for centuries for traditional soy-based food production. A pangenome study using genome data from 462 soybeans worldwide revealed distinct genomic features in Japanese soybeans, as well as several quantitative trait loci alleles that are important for breeding cultivars suited to traditional soy-based foods.

The FarmGTEx Project aims to understand genetic control of gene activity under diverse biological and environmental contexts in domestic animals, providing a foundation for improving animal precision breeding, adaptation and human health.

This Perspective explores causal machine learning in single-cell genomics, addressing challenges such as generalization, interpretability and cell dynamics, while highlighting advances and the potential to uncover new insights into cellular mechanisms.

SF-GWAS is a workflow for secure, federated genome-wide association studies, implementing accurate, privacy-preserving principal-component analysis, linear/logistic regression and linear mixed model methods for biobank-scale multisite analyses.

Genome-wide association analyses identify new risk loci for heart failure and its subtypes, extending understanding of the underlying biological pathways and disease mechanisms.

Common-variant and rare-variant association analyses combining datasets from multiple populations yield insights into the genetic architecture of all-cause heart failure across the allele-frequency spectrum.

Cross-ancestry and sex-stratified genome-wide association analyses for 21 amygdala volumetric traits in East Asian and European individuals provide insight into the genetic architectures of amygdala and subnucleus volumes.

Analysis of whole-genome sequencing data from Iceland and the UK Biobank identifies an excess burden of rare loss-of-function variants in HECTD2 and AKAP11 in individuals diagnosed with bipolar disorder.

The cobraa model extends the pairwise sequentially Markovian coalescent to identify structured population history by examination of the model transition matrix. Applied to human polymorphism data, cobraa identifies an ancient admixture event ancestral to all modern humans.

gLike infers population demographic histories with a variety of complex admixture events by analysis of graphs of states, which conceptualize the relationships of all lineages found in trees encoded in the ancestral recombination graph.

The expansion of clones with distinct SERPINA1 somatic mutants in the livers of alpha-1 antitrypsin deficiency (A1AT) patients is consistent with convergent evolution. These variants interfere with the auto-polymerization and intra-ER accumulation of the Z-A1AT protein, thus highlighting potentially targetable domains.

Whole-genome sequencing and mutational signature analysis of 265 head and neck cancer samples collected from eight different countries provide insight into the vital contribution of tobacco smoke in disease etiology.

NicheCompass is a graph variational autoencoder approach for identifying cellular niches defined by cell–cell communication and other interactions, applicable to a broad variety of spatial multi-omic data.

Multiomics and ecological spatial analysis (MESA) calculates ecodiversity-inspired metrics in spatially resolved omics integrated with single-cell data, enabling the quantitative comparison of tissue states across a range of conditions.

Multi-omic analysis of the adult human prostate identifies spatially resolved cell populations with potential links to prostate carcinogenesis.

Transcriptome-wide m⁶A RNA methylation profile in 162 primary prostate tumors identifies m⁶A association with prognostic clinical features and disease aggression.

Borzoi adapts the Enformer sequence-to-expression model to directly predict RNA-seq coverage, enabling the in-silico analysis of variant effects across multiple layers of gene regulation.

Phosphorylation of KAP1 by DNA-PKcs at enhancers regulated by diverse stimuli prevents association with 7SK small nuclear RNPs and CDK9 SUMOylation, thereby activating the P-TEFb complex and promoting enhancer RNA transcription.

Long-read genome assemblies for seven Japanese, three North American and one primitive Glycine max cultivars highlight gene-level structural variation underlying distinct seed morphology phenotypes.

Chromosome-level genome assemblies of allotetraploid Nicotiana tabacum and its ancestors, along with their transcriptomes, epigenomes and genotype and phenotype data for 5,196 N. tabacum germplasms, provide insights into genome evolution and complex trait regulation.

This study provides a de novo resequencing of 802 Camellia accessions and a genome-wide variation map. Genome-wide association study (GWAS) and metabolite GWAS identify genes responsible for crucial agronomic and flavor traits.

A telomere-to-telomere genome assembly of allohexaploid bread wheat cv. Chinese Spring (CS-IAAS), along with transcriptomic and proteomic data, highlights wheat genome complexity and its structural and functional features.

A telomere-to-telomere genome assembly of Upland cotton (Gossypium hirsutum) accession TM-1 including a set of complete centromeres provides insights into centromere evolution following polyploidization.

A telomere-to-telomere genome assembly of the elite Gossypium hirsutum variety ZM113 provides insights into upland cotton short-season adaptation and identifies a repositioned, satellite repeat-rich centromere.

This study develops family-based genome-wide association study methods that maximize power in homogeneous samples through inclusion of singletons and in diverse samples by using all available parental genotypes.

HDL-L is an extension of the high-definition likelihood method that enables local heritability and genetic correlation analysis with higher accuracy and computational efficiency than LAVA.