NPJ Genomic Medicine最新文献

Precision medicine aims to tailor healthcare by integrating individual genetic, epigenetic, transcriptomic, proteomic, and clinical data, collectively referred to as multi-omic data. However, the scale and complexity of such multi-omics datasets challenge classical computing approaches. Quantum computing, which leverages superposition and entanglement (quantum-level correlations between particles), offers a fundamentally new paradigm for accelerating molecular simulations, biomarker discovery, and high-dimensional data analysis. This review explores the convergence of quantum computing and it's potential to provide unmet needs in precision biomedicine research, with emphasis on applications in diagnostic modeling, multi-omic data integration and drug discovery. We highlight early proof-of-concept studies demonstrating the use of quantum machine learning for disease prediction, quantum algorithms for protein folding, and quantum generative models for novel drug design. Hybrid quantum-classical workflows are also already enabling gene network inference and prioritization of variants of uncertain significance, the latter of which is a major focus of multi-omic research worldwide. Emerging directions include digital twin simulations and real-time clinical decision support powered by quantum models. Looking ahead, the long-term vision for quantum computing in biomedicine involves in silico modeling of entire biological systems to simulate cellular responses to perturbations like drug treatments, enabling clinicians to test therapies in virtual patients before real-world application. Despite these advances, practical implementation remains limited by hardware constraints, qubit decoherence, algorithm scalability, and regulatory barriers. Nonetheless, as quantum hardware evolves and AI-aligned quantum algorithms mature, their integration holds transformative potential. Quantum computing may eventually shorten diagnostic timelines, improve therapeutic precision, and make biomedical innovation more globally accessible. We outline a roadmap for translating these technologies into next-generation precision medicine.

Polygenic embryo screening (PES) is used to screen embryos for their genetic likelihood of developing complex conditions and traits. We surveyed 152 U.S. reproductive endocrinology and infertility specialists (REIs) on their views of PES. While most respondents (97%) were at least slightly familiar with PES, general approval of PES was low (12%), with the majority expressing disapproval (46%) or uncertainty (42%). A majority (58%) believed risks outweigh benefits, while only 16% felt benefits outweigh risks. Most clinicians (85-77%) were very or extremely concerned about low accuracy, confusion over results, false expectations, and eugenics. Nonetheless, when asked to vote on whether PES should be allowed, 44% would vote to allow it, 45% would vote to disallow it, and 10% would abstain from voting. REIs showed more support for PES when used to screen for physical and psychiatric health conditions (59-55% approving) rather than behavioral or physical traits (7-6% approving).

UBR5 encodes an E3 ubiquitin-protein ligase which targets distinct N-terminal residues of proteins for degradation. Heterozygous loss-of-function variants were reported in patients with Autism Spectrum Disorder (ASD) and developmental delay, and recently in a cohort of individuals with neurodevelopmental disorders and variable other features. Here, we report three unrelated individuals with de novo loss-of-function variants in UBR5, presenting with ASD and intellectual disability. We review the literature for other de novo predicted loss-of-function variants in probands with ASD or developmental delay (in total n = 11 variants), providing further evidence that UBR5 haploinsufficiency is associated with ASD and atypical neurodevelopmental trajectories, including developmental delay and intellectual disability.

Leukodystrophies are a diverse group of genetic disorders affecting the central nervous system white matter. Since their initial identification over a century ago, significant advancements have been made in understanding their genetic and clinical profiles. Yet, disease modifying therapies are limited, despite significant clinical impact characterized by progressive neurological decline leading to severe disability and early mortality. This underscores the need for advanced disease models to facilitate the understanding of disease mechanisms and the development of early therapeutic interventions. Stem cells have emerged as a transformative tool in leukodystrophy research, enabling the generation of patient-specific cells otherwise inaccessible for study. We have conducted the first scoping review of stem cell-based disease modeling in leukodystrophies, highlighting recent developments, challenges, and future directions in leveraging these models to enhance our understanding and aid in the development of therapies for these debilitating disorders.

Epstein-Barr virus (EBV) is an oncogenic virus ubiquitous in human populations. CD8 T cells play a crucial role in establishing a strong anti-EBV immune response. Among the various inborn errors of immunity (IEI) showing a restricted vulnerability to EBV, TNFRSF9 (CD137, 4-1BB) deficiency was described in 2019 in patients with chronic EBV viremia and EBV-associated lymphoproliferative diseases. We here investigated a patient with a history of chronic EBV infection and CD137 deficiency who had previously undergone transplantation from her HLA-identical brother. We found that the brother was also a homozygous carrier of the same TNFRSF9 variant, explaining the patient's inability to control EBV after transplantation. Remarkably, during a period of spontaneous clinical improvement and EBV control, we detected two somatic variants in the patient, which resulted in the emergence of two independent revertant CD8 T cell clones that accounted for up to 20% of CD8 T cells in peripheral blood. Using single cell RNA sequencing we demonstrated that both revertant clones originated post-transplant from donor-derived cells. We report here the first described case of a somatic reversion phenomenon in TNFRSF9 deficiency, correlating with clinical improvement and paving the way for future gene therapy strategies for this IEI.

Traditional methods for pharmacogenomics (PGx), like those using disease-specific polygenic risk scores (PRS-Dis), often fail to capture the full heritability of drug response, leading to poor predictions. Direct PGx PRS approaches could improve this, but the scarcity of relevant PGx datasets limits the wide application. To overcome these challenges, we introduce PRS-PGx-TL, a novel transfer learning method. It models large-scale disease summary statistics data alongside individual-level PGx data, leveraging both sources to create more accurate prognostic and predictive polygenic risk scores. In PRS-PGx-TL, we further develop a two-dimensional penalized gradient descent algorithm that starts with weights from disease data and then optimizes them using cross-validation. In simulations and an application to IMPROVE-IT (ClinicalTrials.gov, NCT00202878, September 13, 2005) PGx GWAS data, PRS-PGx-TL significantly enhances prediction accuracy and patient stratification compared to traditional PRS-Dis methods. Our approach shows great promise for advancing precision medicine by using an individual's genetic information to guide treatment decisions more effectively.

Polygenic risk scores (PRSs) are promising tools for genetic risk stratification, but their performance across ancestries remains uncertain. We evaluated 64 published PRSs for eight cardiometabolic traits in 4879 Thai individuals using imputed SNP-array data. Cross-sectional and six-year longitudinal analyses were performed to assess predictive performances. PRSs for type 2 diabetes (T2D) and lipid traits showed the strongest utility, with the best-performing LDL-C and TC scores explaining up to 9.8% and 7.8% of trait variance, respectively. The T2D PRS achieved an area under the curve (AUC) of 0.70 and consistently stratified disease risk over time. In contrast, PRSs for glycemic traits and cardiovascular disease (CVD) had weaker predictive value; notably, the best-performing CVD PRS showed an inverse association with disease risk. Reduced SNP retention and ancestry-related linkage disequilibrium differences contributed to variability. These findings highlight both the potential and current limitations of PRSs in underrepresented Southeast Asian populations.

Impairment of axonal transport has been emphasized as a common feature in a series of neurodegenerative diseases (NDs). Variations in DCTN1 have been reported in NDs such as Parkinson's disease (PD), Perry syndrome (PS) and Amyotrophic lateral sclerosis (ALS). The overall objective of this study was to investigate the contribution of DCTN1 variants in different NDs and to explore the correlation between DCTN1 variants and disease phenotypes. We identified a previously published mutation p.G71E in three unrelated PS families. In the PD cohort, 30 putative deleterious variants (PDVs) were identified in DCTN1. Gene-based burden analysis showed a nominal association between DCTN1 rare PDVs and PD (uncorrected p = 0.042); however, this association did not remain statistically significant after multiple testing correction (FDR-corrected p = 0.084). In the ALS cohort, 10 PDVs were all rare damaging missense variants, and the PDVs were not enriched in ALS patients. Our findings first provide the independent evidence that PDVs in DCTN1 may be a risk factor for PD, but do not support the genetic involvement of DCTN1 in ALS of Asian ancestry.

Nineteen genomic regions have been associated with high-grade serous ovarian cancer (HGSOC). We meta-analyzed >22 million variants for 398,238 women from the Ovarian Cancer Association Consortium (OCAC), UK Biobank (UKBB) and Consortium of Investigators of Modifiers of BRCA1/BRCA2 (CIMBA) to identify novel HGSOC susceptibility loci. Eight novel variants were associated with HGSOC risk. An interesting discovery biologically was TP53 3'-UTR SNP rs78378222-T's association with HGSOC (per-T-allele relative risk (RR) = 1.44, 95% CI:1.28-1.62, P = 1.76 × 10^-9). Polygenic scores (PGS) were developed using OCAC and CIMBA data and trained on FinnGen data. The optimal PGS included 64,518 variants and was associated with an odds ratio of 1.46 (95% CI:1.37-1.54) per standard deviation when validated in the UKBB. This study represents the largest HGSOC GWAS to date - demonstrating that improvements in imputation reference panels and increased sample sizes help to identify HGSOC associated variants that previously went undetected, ultimately improving PGS which can improve personalized HGSOC risk prediction.

While Research Electronic Data Capture (REDCap) is widely adopted in rare disease research, its unconstrained data format often lacks native interoperability with global health standards, limiting secondary use. We developed RareLink, an open-source framework implementing our published ontology-based rare disease common data model. It enables standardised data exchange between REDCap, international registries, and downstream analysis tools by linking Global Alliance for Genomics and Health Phenopackets and Health Level 7 Fast Healthcare Interoperability Resources (FHIR) instances conforming to International Patient Summary and Genomics Reporting profiles. RareLink was developed in three phases across Germany, Canada, South Africa, and Japan for registry and data analysis purposes. We defined a simulated Kabuki syndrome cohort and demonstrated data export to Phenopackets and FHIR. RareLink can enhance the clinical utility of REDCap through its global applicability, supporting equitable rare disease research. Broader adoption and coordination with international entities are thus essential to realise its full potential.