Biology subjects

Wei, W.-Q.

Publications and source records attributed to Wei, W.-Q..

11 recordsLinked to original sources

bioRxiv2026

Assessing the Reliability of LLM-Generated Phenotype-Genotype Associations Through External Validation

BackgroundPhenotype-genotype associations underpin precision medicine by enabling disease prevention, early diagnosis, risk stratification, therapeutic target discovery, and personalized treatment. However, the rapid growth of scientific evidence has made manual curation of these associations increasingly labor-intensive, time-consuming, and incomplete. Large Language Models (LLMs) offer a potential path to scalable genomic generation and synthesis of this knowledge, but their ability to accurately identify phenotype-genotype associations and the extent to which these outputs are supported by established genomic knowledge bases remain unclear. Materials and MethodsFour LLMs, Claude Sonnet 4.6, DeepSeek V4 Flash, Gemini 3 Flash Preview, and GPT-5.5, were benchmarked on six zero-shot task categories covering forward and reverse phenotype-gene and phenotype-SNP generation. A total of 4,196 associations were identified from curated inputs and evaluated through a multistage external verification pipeline comprising phenotype normalization, ontology mapping, genomic identifier validation against Ensembl, and evidence verification using both the GWAS Catalog and OMIM. Associations were assigned a fused evidence level of strong, moderate, weak, or none. ResultsOverall, 74.19% of generated associations were matched to at least one external genomic knowledge base; 9.15% received strong support and 54.46% moderate support. Phenotype-gene associations were more verifiable than phenotype-SNP associations (strong or moderate: 67.19% vs 54.06%). Among existing associations, Claude Sonnet 4.6 achieved the highest overall strong or moderate rate (69.2%), followed by GPT-5.5 (65.1%), DeepSeek V4 Flash (61.7%), and Gemini 3 Flash Preview (56.9%). ConclusionLLMs can support scalable generation of candidate phenotype-genotype associations. Performance varied substantially by relation type and was lower for SNP-level and rare disease associations, highlighting both the limitations of current genomic resources and the need for rigorous validation pipelines.

bioinformatics↗

bioRxiv2026

High glucose confers senescence resistance via GLUT1 epigenetic rewiring to blunt immunotherapy responses in esophageal squamous cell carcinoma

Therapeutic resistance and undefined predictive biomarkers severely hinder the clinical popularization of immunotherapy in esophageal squamous cell carcinoma (ESCC). Herein, we identify the glucose transporter 1 (GLUT1) as a critical determinant of immunotherapy resistance. Elevated expression of GLUT1 correlates with poor immunotherapy response and unfavorable prognosis in ESCC patients. GLUT1 deletion or inhibition enhances CD8 T cell infiltration and cytotoxicity, and sensitizes ESCC tumors to anti-PD-1 (-PD1) therapy. Importantly, dietary glucose restriction exhibits equivalent antitumor efficacy to GLUT1 inhibition when combined with -PD1. Mechanistically, GLUT1 establishes a positive feedback loop with HAT1 and FOXM1, which epigenetically remodels chromatin accessibility to suppress tumor cell senescence, thereby impeding CD8 T cell-mediated antitumor immunity. Our findings highlight GLUT1 as a predictive biomarker of immunotherapy resistance and suggest dietary glucose restriction as a viable strategy to potentiate immunotherapy efficacy in ESCC. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=188 HEIGHT=200 SRC="FIGDIR/small/738372v1_ufig1.gif" ALT="Figure 1"> View larger version (52K): org.highwire.dtl.DTLVardef@c2c5d5org.highwire.dtl.DTLVardef@14a34c7org.highwire.dtl.DTLVardef@cf05e4org.highwire.dtl.DTLVardef@18bc354_HPS_FORMAT_FIGEXP M_FIG C_FIG SHORT SUMMARYDong et al. identify glucose transporter 1 (GLUT1) as a predictor of poor response to immunotherapy in esophageal squamous cell carcinoma. GLUT1 promotes immune evasion by forming a positive feedback loop with the HAT1/FOXM1 anti-senescence axis, thereby epigenetically remodeling chromatin accessibility. GLUT1 inhibition or dietary glucose restriction restores CD8 T-cell-mediated antitumor immunity and improves the efficacy of anti-PD-1 therapy in preclinical models.

cancer biology↗

bioRxiv2024

Single-Cell Transcriptomic Analysis Highlights Specific Cell Types of Wheat Manipulated by Fusarium graminearum Leading to Susceptibility

Phytopathogenic fungi can be recognized by the plant immune system and trigger host defenses, but adapted pathogens cause susceptibility. How different cell types cooperate and orchestrate biological processes in response to heterogeneous colonization of organs by adapted and non-adapted pathogens remains largely unknown. Here we employed single-cell RNA sequencing to dissect the responses of wheat coleoptiles to infection by the adapted fungal pathogen Fusarium graminearum (Fgr) and the non-adapted fungal pathogen Fusarium oxysporum f. sp. cubense (Foc) at 1-, 2-, and 3-days post-inoculation. We profiled the transcriptomes of over 90,000 cells and identified eight major cell types in coleoptiles: stomata, epidermis, chlorenchyma, parenchyma, outer sheath, inner sheath, phloem, and procambium. Differential expression analyses showed that the capacity of different cell types to respond to fungal infection varied. The upregulation of immune pathways was compartmentalized in nonhost resistance to Foc, but widespread in susceptible interaction with Fgr. Pseudotime analyses revealed continuous cell state transitions in the disease progression of infected cell types. Our work indicates that the phloem and outer sheath are specific cell types that collaborate for the rapid onset of nonhost resistance. Fgr induces a state of low transcriptional activity in the chlorenchyma. Cell trajectory analysis suggests that the competition between immune and susceptible processes in parenchyma results in specific cell states that are favored by the adapted pathogen Fgr. Overall, this work explains how cell types collaborate and are manipulated during fungal infections, providing insight into the intercellular mechanisms of plant immunity.

plant biology↗

bioRxiv2021

Projecting genetic associations through gene expression patterns highlights disease etiology and drug mechanisms

Genes act in concert with each other in specific contexts to perform their functions. Determining how these genes influence complex traits requires a mechanistic understanding of expression regulation across different conditions. It has been shown that this insight is critical for developing new therapies. In this regard, the role of individual genes in disease-relevant mechanisms can be hypothesized with transcriptome-wide association studies (TWAS), which have represented a significant step forward in testing the mediating role of gene expression in GWAS associations. However, modern models of the architecture of complex traits predict that gene-gene interactions play a crucial role in disease origin and progression. Here we introduce PhenoPLIER, a computational approach that maps gene-trait associations and pharmacological perturbation data into a common latent representation for a joint analysis. This representation is based on modules of genes with similar expression patterns across the same conditions. We observed that diseases were significantly associated with gene modules expressed in relevant cell types, and our approach was accurate in predicting known drug-disease pairs and inferring mechanisms of action. Furthermore, using a CRISPR screen to analyze lipid regulation, we found that functionally important players lacked TWAS associations but were prioritized in trait-associated modules by PhenoPLIER. By incorporating groups of co-expressed genes, PhenoPLIER can contextualize genetic associations and reveal potential targets missed by single-gene strategies.

bioinformatics↗

bioRxiv2021

Neptune: An environment for the delivery of genomic medicine

PurposeGenomic medicine holds great promise for improving healthcare, but integrating searchable and actionable genetic data into electronic health records remains a challenge. Here, we describe Neptune, a system for managing the interaction between a clinical laboratory and an electronic health record system. MethodsWe developed Neptune and applied it to two clinical sequencing projects that required report customization, variant reanalysis and EHR integration. ResultsNeptune enabled the analysis of data for generation of and delivery to EHR systems of over 15,000 clinical genomic reports. These projects demanded customizable clinical reports that contained a variety of genetic data types including SNVs, CNVs, pharmacogenomics and polygenic risk scores. Two variant reanalysis activities were also supported, highlighting this important workflow. ConclusionsMethods are needed for delivering structured genetic data to EHRs. This need extends beyond developing data formats to providing infrastructure that manages the reporting process itself. Neptune was successfully applied on two high-throughput clinical sequencing projects to build and deliver clinical reports to EHR systems. The software is open and available at https://gitlab.com/bcm-hgsc/neptune.

bioinformatics↗

bioRxiv2019

A Phenome-wide Mendelian Randomisation study on genetically determined serum urate levels in UK Biobank cohort

IntroductionThe role of serum urate level has been extensively investigated in observational studies. However, the extent of any causal effect remains unclear, making it difficult to evaluate its clinical relevance.\n\nObjectivesTo explore any causal or pleiotropic association between serum urate level and a broad spectrum of disease outcomes.\n\nMethodsPhenome-wide association study (PheWAS) together with a Bayesian analysis of tree-structured phenotypic models (TreeWAS) was performed to examine disease outcomes related to genetically determined serum urate levels in 339,256 UK Biobank participants. Mendelian Randomisation (MR) analyses were performed to replicate significant findings using various GWAS consortia data. Sensitivity analyses were conducted to examine possible pleiotropic effects on metabolic traits of the genetic variants used as instruments for serum urate.\n\nResultsPheWAS analysis, examining the association with 1,431 disease outcomes, identified a multitude of disease outcomes including gout, hypertension, hypercholesterolemia, and heart diseases (e.g., coronary atherosclerosis, myocardial infarction, and ischaemic heart disease) that were associated (p<3.35e-04) with genetically determined serum urate levels. TreeWAS analysis, examining 10,750 ICD-10 diagnostic terms, identified more sub-phenotypes of cardiovascular and cerebrovascular diseases (e.g., angina pectoris, heart failure, cerebral infarction). MR analysis successfully replicated the association with gout, hypertension, heart diseases and blood lipid levels, but indicated the existence of genetic pleiotropy. Sensitivity analyses support an inference that pleiotropic effects of genetic variants on urate and metabolic traits contribute to the observational associations with cardiovascular diseases.\n\nConclusionsHigh serum urate levels are associated with different types of cardiac events. The finding of genetic pleiotropy indicates the existence of common upstream pathological elements influencing both urate and metabolic traits, and this may suggest new opportunities and challenges for developing drugs targeting a more distal mediator that would be beneficial for both the treatment of gout and the prevention of cardiovascular comorbidities.

epidemiology↗

bioRxiv2019

A missense variant in Mitochondrial Amidoxime Reducing Component 1 gene and protection against liver disease

Analyzing 5770 all-cause cirrhosis cases and 572,850 controls from seven cohorts, we identify a missense variant in the Mitochondrial Amidoxime Reducing Component 1 gene (MARC1 p.A165T) that associates with protection from all-cause cirrhosis (OR 0.88, p=2.1*10-8). This same variant also associates with lower levels of hepatic fat on computed tomographic imaging and lower odds of physician-diagnosed fatty liver as well as lower blood levels of alanine transaminase (-0.012 SD, 1.4*10-8), alkaline phosphatase (-0.019 SD, 6.6*10-9), total cholesterol (-0.037 SD, p=1*10-18) and LDL cholesterol (-0.035 SD, p=7.3*10-16). Carriers of rare protein-truncating variants in MARC1 had lower liver enzyme levels, cholesterol levels, and reduced odds of liver disease (OR 0.19, p= 0.04) suggesting that deficiency of the MARC1 enzyme protects against cirrhosis.

genomics↗

bioRxiv2018

Genetic variants related to antihypertensive targets inform drug efficacy and side effects

BackgroundDrug effects can be investigated through natural variation in the genes for their protein targets. We aimed to use this approach to explore the potential side effects and repurposing potential of antihypertensive drugs, which are amongst the most commonly used medications worldwide.\n\nMethodsWe identified genetic instruments for antihypertensive drug classes as variants in the gene for the corresponding target that associated with systolic blood pressure at genome-wide significance. To validate the instruments, we compared Mendelian randomisation (MR) estimates for drug effects on coronary heart disease (CHD) and stroke risk to randomised controlled trial (RCT) results. Phenome-wide association study (PheWAS) in the UK Biobank was performed to identify potential side effects and repurposing opportunities, with findings investigated in the Vanderbilt University Biobank (BioVU) and in observational analysis of the UK Biobank.\n\nFindingsWe identified suitable genetic instruments for beta-blockers (BBs) and calcium channel blockers (CCBs). MR estimates for their effect on CHD and stroke risk respectively were comparable to results from RCTs against placebo. PheWAS in the UK Biobank identified an association of the CCB genetic risk score (scaled to drug effect) with increased risk of diverticulosis (odds ratio [OR] 1.23, 95%CI 1.10-1.38), with a consistent estimate found in BioVU (OR 1.16, 95%CI 0.94-1.44). Association with diverticulosis was further supported in observational analysis of CCB use in the UK Biobank (OR 1.08, 95%CI 1.02-1.15).\n\nInterpretationWe identified valid genetic instruments for BBs and CCBs. Using genetic and observational approaches, we highlighted a previously unreported potential detrimental effect of CCBs on risk of diverticulosis. This work serves as a proof of concept that investigation of genetic variants can offer a complementary approach to exploring the efficacy and side effects of anti-hypertensive medications.\n\nFundingWellcome Trust.

genetics↗

bioRxiv2018

Developing and Evaluating Mappings of ICD-10 and ICD-10-CM codes to Phecodes

BackgroundThe PheCode system was built upon the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) for phenome-wide association studies (PheWAS) in the electronic health record (EHR).\n\nObjectiveHere, we present our work on the development and evaluation of maps from ICD-10 and ICD-10-CM codes to PheCodes.\n\nMethodsWe mapped ICD-10 and ICD-10-CM codes to PheCodes using a number of methods and resources, such as concept relationships and explicit mappings from the Unified Medical Language System (UMLS), Observational Health Data Sciences and Informatics (OHDSI), Systematized Nomenclature of Medicine - Clinical Terms (SNOMED CT), and National Library of Medicine (NLM). We assessed the coverage of the maps in two databases: Vanderbilt University Medical Center (VUMC) using ICD-10-CM and the UK Biobank (UKBB) using ICD-10. We assessed the fidelity of the ICD-10-CM map in comparison to the gold-standard ICD-9-CM[->]PheCode map by investigating phenotype reproducibility and conducting a PheWAS.\n\nResultsWe mapped >75% of ICD-10-CM and ICD-10 codes to PheCodes. Of the unique codes observed in the VUMC (ICD-10-CM) and UKBB (ICD-10) cohorts, >90% were mapped to PheCodes. We observed 70-75% reproducibility for chronic diseases and <10% for an acute disease. A PheWAS with a lipoprotein(a) (LPA) genetic variant, rs10455872, using the ICD-9-CM and ICD-10-CM maps replicated two genotype-phenotype associations with similar effect sizes: coronary atherosclerosis (ICD-9-CM: P < .001, OR = 1.60 vs. ICD-10-CM: P < .001, OR = 1.60) and with chronic ischemic heart disease (ICD-9-CM: P < .001, OR = 1.5 vs. ICD-10-CM: P < .001, OR = 1.47).\n\nConclusionsThis study introduces the initial \"beta\" versions of ICD-10 and ICD-10-CM to PheCode maps that will enable researchers to leverage accumulated ICD-10 and ICD-10-CM data for high-throughput PheWAS in the EHR.

bioinformatics↗

bioRxiv2018

Learning from Longitudinal Data in Electronic Health Record and Genetic Data to Improve Cardiovascular Event Prediction

BackgroundCurrent approaches to predicting Cardiovascular disease rely on conventional risk factors and cross-sectional data. In this study, we asked whether: i) machine learning and deep learning models with longitudinal EHR information can improve the prediction of 10-year CVD risk, and ii) incorporating genetic data can add values to predictability.\n\nMethodsWe conducted two experiments. In the first experiment, we modeled longitudinal EHR data with aggregated features and temporal features. We applied logistic regression (LR), random forests (RF) and gradient boosting trees (GBT) and Convolutional Neural Networks (CNN) and Recurrent Neural Networks, using Long Short-Term Memory (LSTM) units. In the second experiment, we proposed a late-fusion framework to incorporate genetic features.\n\nResultsOur study cohort included 109, 490 individuals (9,824 were cases and 99, 666 were controls) from Vanderbilt University Medical Centers (VUMC) de-identified EHRs. American College of Cardiology and the American Heart Association (ACC/AHA) Pooled Cohort Risk Equations had areas under receiver operating characteristic curves (AUROC) of 0.732 and areas under receiver under precision and recall curves (AUPRC) of 0.187. LSTM, CNN and GBT with temporal features achieved best results, which had AUROC of 0.789, 0.790, and 0.791, and AUPRC of 0.282, 0.280 and 0.285, respectively. The late fusion approach achieved a significant improvement for the prediction performance.\n\nConclusionsMachine learning and deep learning with longitudinal features improved the 10-year CVD risk prediction. Incorporating genetic features further enhanced 10-year CVD prediction performance, underscoring the importance of integrating relevant genetic data whenever available in the context of routine care.

epidemiology↗

bioRxiv2018

Using Topic Modeling via Non-negative Matrix Factorization to Identify Relationships between Genetic Variants and Disease Phenotypes: A Case Study of Lipoprotein(a) (LPA)

Genome-wide and phenome-wide association studies are commonly used to identify important relationships between genetic variants and phenotypes. Most of these studies have treated diseases as independent variables and suffered from heavy multiple adjustment burdens due to the large number of genetic variants and disease phenotypes. In this study, we propose using topic modeling via non-negative matrix factorization (NMF) for identifying associations between disease phenotypes and genetic variants. Topic modeling is an unsupervised machine learning approach that can be used to learn the semantic patterns from electronic health record data. We chose rs10455872 in LPA as the predictor since it has been shown to be associated with increased risk of hyperlipidemia and cardiovascular diseases (CVD). Using data of 12,759 individuals from the biobank at Vanderbilt University Medical Center, we trained a topic model using NMF from 1,853 distinct phecodes extracted from the cohorts electronic health records and generated six topics. We quantified their associations with rs10455872 in LPA. Topics indicating CVD had positive correlations with rs10455872 (P < 0.001), replicating a previous finding. We also identified a negative correlation between LPA and a topic representing lung cancer (P < 0.001). Our results demonstrate the applicability of topic modeling in exploring the relationship between the genome and clinical diseases.\n\nAuthor summaryIdentifying the clinical associations of genetic variants remains crucial in understanding how the human genome modulates disease risk. Traditional phenome-wide association studies consider each disease phenotype as an independent variable, however, diseases often present as complex clusters of comorbid conditions. In this study, we propose using topic modeling to model electronic health record data as a mixture of topics (e.g., disease clusters or relevant comorbidities) and testing associations between topics and genetic variants. Our results demonstrated the feasibility of using topic modeling to replicate and discover novel associations between the human genome and clinical diseases.

genetics↗