196 papers
This paper introduces DAVP, a hierarchical agentic AI system that significantly outperforms expert clinicians in speed and accuracy for genetic diagnosis by dynamically prioritizing variants through a specialized three-component workflow that integrates gene pre-screening, semantic knowledge graph analysis, and in-context learning to address the complexities of rare disease detection.
This study demonstrates that short tandem repeats (STRs) significantly contribute to the genetic architecture of age-related hearing loss, particularly explaining a larger portion of the heritability for metabolic phenotypes than sensory ones, while identifying specific STR variants and rare repeat burdens associated with increased metabolic risk and reduced sensory risk.
This study analyzing 490,640 UK Biobank genomes found that severe combined immunodeficiency (SCID) variants exhibit high penetrance and a low prevalence of biallelic pathogenic variants, supporting the inclusion of SCID in genomic newborn screening while highlighting the need for careful reporting of hypomorphic variants.
This study developed and cross-tissue validated a novel machine learning-based methylation profile score using whole blood data that successfully predicts individual salivary cortisol responses to stress, revealing associations with immune and stress-related pathways and offering a potential epigenetic biomarker for improving clinical risk assessment of HPA axis dysregulation.
In a large observational cohort study of European ancestry individuals, higher polygenic scores for major depressive disorder were significantly associated with increased healthcare utilization and comorbidity burden, both in those without a clinical diagnosis and those with a confirmed diagnosis, suggesting that these scores may serve as valuable biomarkers for predicting health outcomes in real-world settings.
This study characterizes the genetic architecture of the French OFSEP-HD Multiple Sclerosis cohort, revealing significant North-African ancestry among patients who did not self-report such origins and providing an open synthetic dataset to facilitate future research.
This study utilizes multi-omics and machine learning approaches to uncover sex-specific genetic links between obesity, type 2 diabetes, and Alzheimer's disease, identifying shared biological pathways and repurposing the heart failure drug levosimendan as a potential candidate for Alzheimer's treatment.
This study characterizes the somatic and transcriptional profiles of breast tumors in 748 Hispanic/Latina women, revealing similarities to non-Hispanic White women overall but identifying distinct features such as increased CTCF mutations, a higher prevalence of favorable immune ecotypes linked to Indigenous American ancestry, and a germline APOBEC3A/B deletion associated with specific mutational signatures.
This study analyzed whole-exome sequencing data from 6,739 Russian individuals to identify 75 pathogenic or likely pathogenic variants in 18 genes associated with monogenic syndromes, demonstrating that 1.77% of the population carries clinically actionable genetic risks for adverse drug reactions that could be mitigated through personalized pharmacotherapy.
This study utilizes integrative transcriptomic analysis of postmortem human brain tissue to reveal that long noncoding RNAs (lncRNAs) are extensively dysregulated in the reward and executive circuits of opioid use disorder, exhibiting cell-type specificity and circadian disruption that contribute to the disease's underlying pathology.
This study presents a comprehensive network-based atlas of human skeletal muscle aging, integrating deep transcriptomic profiles from 1,675 biopsies with advanced spatial and single-cell technologies to construct quantitative network models that reveal cell-specific aging mechanisms, distinct exercise responses, and a novel transcriptomic age clock.
This study demonstrates that pervasive non-causal "tagging" effects, driven by shared transcription factor binding sites and peak co-accessibility, confound enhancer-gene linking in single-cell multiome data, necessitating the use of fine-mapping methods like SuSiE to distinguish true causal regulatory relationships from spurious correlations.
This retrospective analysis of 6,102 Russian exomes establishes a comprehensive reference for pharmacogenomic variant and HLA allele frequencies, confirming the utility of whole-exome sequencing for population screening while highlighting its critical limitations in detecting non-coding variants and accurately determining CYP2D6 copy numbers.
This study presents the first genome-wide association analysis of lateral epicondylopathy, identifying two significant risk loci that implicate SIBLING genes in extracellular matrix remodelling and challenge previously reported collagen gene associations.
This study utilized whole exome sequencing on 117 patients with suspected monogenic cerebral small vessel disease to identify novel gene associations, including a significant burden of variants in ABCC6 and new links to genes such as MYH11, NOTCH1, and seven others, thereby highlighting the need for expanded genetic screening and functional characterization to improve diagnosis and understanding of CSVD pathogenesis.