SEARCH · Search PubMed

Results for “Phenomics”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 145 records · Page 8Linked to original sources

PubMedDate not supplied

DNA microarrays: experimental issues, data analysis, and application to bacterial systems.

DNA microarrays are currently used to study the transcriptional response of many organisms to genetic and environmental perturbations. Although there is much room for improvement of this technology, its potential has been clearly demonstrated in the past 5 years. The general consensus is that the bottleneck is now located in the processing and analysis of transcriptome data and its use for purposes other than the quantification of changes in gene expression levels. In this article we discuss technological aspects of DNA microarrays, statistical and biological issues pertinent to the design of microarray experiments, and statistical tools for microarray data analysis. A review on applications of DNA microarrays in the study of bacterial systems is presented. Special attention is given to studies in the following areas: (1) bacterial response to environmental changes; (2) gene identification, genome organization, and transcriptional regulation; and (3) genetic and metabolic engineering. Soon, the use of DNA microarray technologies in conjunction with other genome/system-wide analyses (e.g., proteomics, metabolomics, fluxomics, phenomics, etc.) will provide a better assessment of genotype-phenotype relationships in bacteria, which serve as a basis for understanding similar processes in more complex organisms.

Algorithms↗

PubMed2002

DNA microarrays for functional plant genomics.

DNA microarray technology is a key element in today's functional genomics toolbox. The power of the method lies in miniaturization, automation and parallelism permitting large-scale and genome-wide acquisition of quantitative biological information from multiple samples. DNA microarrays are currently fabricated and assayed by two main approaches involving either in situ synthesis of oligonucleotides ('oligonucleotide microarrays') or deposition of pre-synthesized DNA fragments ('cDNA microarrays') on solid surfaces. To date, the main applications of microarrays are in comprehensive, simultaneous gene expression monitoring and in DNA variation analyses for the identification and genotyping of mutations and polymorphisms. Already at a relatively early stage of its application in plant science, microarrays are being utilized to examine a range of biological issues including the circadian clock, plant defence, environmental stress responses, fruit ripening, phytochrome A signalling, seed development and nitrate assimilation. Novel insights are obtained into the molecular mechanisms co-ordinating metabolic pathways, regulatory and signalling networks. Exciting new information will be gained in the years to come not only from genome-wide expression analyses on a few model plant species, but also from extensive studies of less thoroughly studied species on a more limited scale. The value of microarray technology to our understanding of living processes will depend both on the amount of data to be generated and on its clever exploration and integration with other biological knowledge arising from complementary functional genomics tools for 'profiling' the genome, proteome, metabolome and phenome.

Gene Expression Profiling↗

PubMed2002

Behavioural and physiological characterization of inbred mouse strains: prospects for elucidating the molecular mechanisms of mammalian learning and memory.

With the advent of recombinant DNA methodology, it has become possible to dissect the molecular mechanisms of complex traits, including brain function and behaviour. The increasing amount of available information on the genomes of mammalian organisms, including our own, has facilitated this research. The present review focuses on a somewhat neglected area of genetics, one that involves the study of inbred mouse strains. It is argued that the use of inbred mice is complementary to transgenic approaches in the analysis of molecular mechanisms of complex traits. Whereas transgenic technology allows one to manipulate a single gene and investigate the in vivo effects of highly specific, artificially induced mutations, the study of inbred mouse strains should shed light on the roles of naturally occurring allelic variants in brain function and behaviour. Systematic characterization of the behavioural, electrophysiological, neurochemical, and neuroanatomical properties of a large number of inbred strains is required to elucidate mechanisms of mammalian brain function and behaviour. In essence, a 'mouse phenome' project is needed, entailing the construction of databases to investigate possible causal relationships amongst the phenotypical characteristics. This review focuses on electrophysiological and behavioural characterization of mouse strains. Nevertheless, it is emphasized that the full potential of the analysis of inbred mouse strains may be attained if techniques of numerous disciplines, including gene expression profiling, biochemical analysis, and quantitative trait loci (QTL) mapping, to name but a few, are also included.

Animals↗

PubMed2003

A rating scale for wildness and ease of handling laboratory mice: results for 21 inbred strains tested in two laboratories.

Rating scales for difficulty in capturing and holding mice were devised that proved to be easy to use and highly sensitive to differences among mouse strains on the A and B priority lists of the Mouse Phenome Project. The simplicity of the scales makes it feasible to rate wildness during behavioral test sessions without adding much to testing time or distracting the technician from the principal task at hand. Overall wildness and placidity ratings obtained by combining capture and hold ratings provide a good impression of the difficulties encountered while working with lab mice in the course of complex experiments. Ratings of 21 inbred strains during the course of 15 behavioral tests in two laboratories demonstrated that the SPRET/Ei, PERA/Ei, CAST/Ei and SWR/J strains were particularly difficult to handle. The NOD/LtJ strain posed no special challenge in the Edmonton laboratory but was very difficult to handle in the Portland lab. The rating scales should be useful for judging the difficulties in working with novel targeted or induced mutations in mice as well as effects of a variety of environmental treatments or drugs.

Animals↗

PubMed2024

Contribution of copy number variations to education, socioeconomic status and cognition from a genome-wide study of 305,401 subjects.

Educational attainment (EA), socioeconomic status (SES) and cognition are phenotypically and genetically linked to health outcomes. However, the role of copy number variations (CNVs) in influencing EA/SES/cognition remains unclear. Using a large-scale (n = 305,401) genome-wide CNV-level association analysis, we discovered 33 CNV loci significantly associated with EA/SES/cognition, 20 of which were novel (deletions at 2p22.2, 2p16.2, 2p12, 3p25.3, 4p15.2, 5p15.33, 5q21.1, 8p21.3, 9p21.1, 11p14.3, 13q12.13, 17q21.31, and 20q13.33, as well as duplications at 3q12.2, 3q23, 7p22.3, 8p23.1, 8p23.2, 17q12 (105 kb), and 19q13.32). The genes identified in gene-level tests were enriched in biological pathways such as neurodegeneration, telomere maintenance and axon guidance. Phenome-wide association studies further identified novel associations of EA/SES/cognition-associated CNVs with mental and physical diseases, such as 6q27 duplication with upper respiratory disease and 17q12 (105 kb) duplication with mood disorders. Our findings provide a genome-wide CNV profile for EA/SES/cognition and bridge their connections to health. The expanded candidate CNVs database and the residing genes would be a valuable resource for future studies aimed at uncovering the biological mechanisms underlying cognitive function and related clinical phenotypes.

Humans↗

PubMed2024

Maternal COVID-19 infection associated with offspring neurodevelopmental disorders.

Maternal COVID-19 infection increases the incidence of neurodevelopmental disorders (NDDs) in offspring, although the underlying mechanisms have not been elucidated. This study demonstrated that COVID-19 infection during pregnancy disrupted the balance of maternal and fetal immune environments, driving alterations in astrocytes, endothelial cells, and excitatory neurons. A risk score was established using 47 unique genes in the single-cell transcriptome of gestational mothers. The high risk score in CD4 proliferating T cell level served as an indicator for increased risk of offspring NDDs. Summary-based Mendelian randomization and phenome-wide association study analyses were conducted to identify the causal association of the transcriptional changes with the increased risk of offspring NDDs. Additionally, 10 drugs were identified as potential therapeutic candidates. Our findings support a model where the maternal COVID-19 infection changed the levels of CD4 proliferating T cells, leading to the alterations of astrocytes, endothelial cells, and excitatory neurons in offspring, contributing to the increased risk of NDDs in these individuals.

Humans↗

PubMed2026

Immune cell-specific genetic architecture of Alzheimer's disease revealed by multi-omics analysis for therapeutic target discovery and prioritization.

Alzheimer's disease (AD) is a multifactorial neurodegenerative condition in which accumulating genetic and molecular evidence implicates dysregulation of peripheral immune processes in disease pathogenesis. Nevertheless, the contribution of distinct peripheral immune cell subsets and associated gene regulatory landscapes to AD risk remains incompletely defined. To address this gap, we integrated single-cell expression quantitative trait loci (sc‑eQTL) data from the OneK1K cohort with AD GWAS summary statistics. We systematically interrogated immune cell-specific genes for their contributions to AD risk by integrating genetic causal inference with Bayesian colocalization analyses, and identified 24 eGenes that passed both the MR significance threshold (P < 0.05) and the criterion for strong shared genetic signals (PP.H4 > 0.8). Notable candidates included GATS, HLA-DOB, HLA-DQA1, PM20D1, and others, with each gene demonstrating a cell-type-specific association restricted to its corresponding immune cell type, such as monocytes, CD8 + T cells, or B cells. Independent peripheral blood single-cell transcriptomic data further supported disease-associated shifts in cell-type-specific expression patterns in AD. Phenome-wide association studies (PheWAS) indicated limited associations with off-target traits, indicating a favorable safety profile for therapeutic intervention, with the exceptions of B4GALNT3, PM20D1, and CNN2. Integration of immune gene targets with pharmacological databases yielded three candidate compound, including NSC321521 (targeting HLA-DQA1), phenoxybenzamine (targeting GSTP1), and rimexolone (targeting BIN1). Among these compounds, Predicted blood-brain barrier permeability was observed only for phenoxybenzamine and rimexolone, with docking studies indicating stable interactions, such as those between NSC321521 and HLA-DQA1, phenoxybenzamine and GSTP1, and rimexolone and BIN1. This integrative approach highlights key immune‑cell‑specific genes involved in AD and proposes repurposable drugs with central nervous system potential, paving the way for more targeted immunomodulatory strategies in AD.

Humans↗

PubMed2025

Genome-wide association study meta-analysis provides insights into the etiology of heart failure and its subtypes.

Heart failure (HF) is a major contributor to global morbidity and mortality. While distinct clinical subtypes, defined by etiology and left ventricular ejection fraction, are well recognized, their genetic determinants remain inadequately understood. In this study, we report a genome-wide association study of HF and its subtypes in a sample of 1.9 million individuals. A total of 153,174 individuals had HF, of whom 44,012 had a nonischemic etiology (ni-HF). A subset of patients with ni-HF were stratified based on left ventricular systolic function, where data were available, identifying 5,406 individuals with reduced ejection fraction and 3,841 with preserved ejection fraction. We identify 66 genetic loci associated with HF and its subtypes, 37 of which have not previously been reported. Using functionally informed gene prioritization methods, we predict effector genes for each identified locus, and map these to etiologic disease clusters through phenome-wide association analysis, network analysis and colocalization. Through heritability enrichment analysis, we highlight the role of extracardiac tissues in disease etiology. We then examine the differential associations of upstream risk factors with HF subtypes using Mendelian randomization. These findings extend our understanding of the mechanisms underlying HF etiology and may inform future approaches to prevention and treatment.

Humans↗

PubMed2026

Unsupervised characterization of 100,272 EHR patients identifies high-risk groups and comorbidities linked to premature aging.

Electronic health records (EHRs) contain extensive multidimensional patient data, presenting challenges for the discovery of novel and meaningful clinical patterns. Unsupervised clustering of high-dimensional clinical data holds great potential for identifying novel clinical patterns. Here, we performed unsupervised clustering and characterized 100,272 patients in the Electronic Medical Records and GEnomics (eMERGE) Network. We identified 70 clusters defined by distinct comorbidity patterns. Meanwhile, age and sex are also strongly associated with patient stratification, influencing phenotype prevalence and onset time. Notably, phenotype onset time accurately predicted chronological age and was significantly associated with overall mortality risk. Besides age and sex, we assessed the contribution of genetic variation to phenotype development and observed evidence of cross-phenotype associations influencing cluster membership and comorbidity patterns. However, the role of genetics recedes during aging. We also identified several high-risk clusters with elevated Charlson Comorbidity Index (CCI) scores and validated these findings in an independent cohort. Further analysis of these clusters revealed phenotypes linked to premature aging and highlighted a survival selection among older participants in observational studies. Overall, this study enables phenome-wide unsupervised patient stratification for multimorbidity discovery in largely unannotated clinical data, offering valuable insights into patient stratification, comorbidity analysis, aging, and health outcomes.

Journal Article↗

PubMed2026

GFPT1 as a cross-ancestry validated target for degenerative spinal disease: genetic association in a Chinese cohort and functional characterization in zebrafish.

Degenerative spinal disease (DSD), including spinal stenosis and spondylosis, lacks effective pharmacological treatment. To identify druggable targets and assess cross-ancestry applicability, we integrate multi-omics analyses using Summary-data-based Mendelian Randomization (SMR), colocalization, and two-sample Mendelian randomization with European whole-blood, peripheral-blood, and CSF eQTL/pQTL datasets, followed by whole-genome sequencing (WGS) validation in a Chinese cohort. We identify 7 genes/proteins associated with spinal stenosis and 5 with spondylosis, with GFPT1, GPX1, and SERPINA1 shared by both. Two-sample MR further supports the causal associations of these targets with DSD. Phenome-wide MR prioritization selects GFPT1 and GPX1 as favorable candidates with no predicted adverse effects and potential beneficial effects on hypertension. In the Chinese cohort (67 lumbar spinal stenosis patients and 100 controls), WGS identifies 4 GFPT1 cis-eQTL loci (rs13016371, rs35392088, rs12997521, and rs13019789) associated with lumbar spinal stenosis risk; all risk alleles are linked to increased GFPT1 expression, and all 24 variant carriers show L4/L5 stenosis on imaging. Druggability analysis identifies IOX1 as the sole preclinical-stage compound targeting GFPT1, and molecular docking supports robust binding to GFPT1 (- 6.39 kcal/mol). Functional assays show that IOX1 directly inhibits GFPT1 enzymatic activity and induces fructose-6-phosphate accumulation. In zebrafish, IOX1 significantly rescues GFPT1-induced degenerative phenotypes. These findings establish GFPT1 as a cross-ancestry validated therapeutic target for DSD and nominate IOX1 as a promising disease-modifying candidate.

Animals↗

PubMed2026

Genetic evidence supports the combined targeting of lipoprotein(a) and LDL cholesterol to reduce coronary artery disease risk.

Distinct genetic mechanisms govern how lipoprotein(a) (Lp(a)) and low-density lipoprotein cholesterol (LDL-C) promote atherosclerosis. It remains unclear whether targeting both provides additive cardiovascular benefits. Here we use coding loss-of-function variants in LPA and PCSK9 and genetic scores associated with Lp(a) and LDL-C levels to evaluate the effects of lowering Lp(a) and LDL-C on coronary artery disease (CAD) risk. Among 408,039 individuals from the UK Biobank, LPA or PCSK9 loss-of-function carriers have lower CAD risk than noncarriers (odds ratio (OR) 0.91 and 0.81). Carriers of both variants have even lower CAD risk (OR 0.73). Genetic lowering of Lp(a) and LDL-C showed a stronger reduction of CAD risk (OR 0.70) than either trait individually (OR 0.85 and 0.81) in the two-factor genetic score analysis. Among statin users, Lp(a) reduction was linearly associated with CAD risk. A phenome-wide association study revealed that combined therapy was associated with cardiometabolic benefits without adverse effects. The additive benefits were replicated in 65,171 individuals from the Mass General Brigham Biobank.

Humans↗

PubMed2006

Ab initio calculations on SCl2 and low-lying cationic states of SCl2+: Franck-Condon simulation of the UV photoelectron spectrum of SCl2.

Geometry optimization calculations were carried out on the (approximate)X (1)A(1) state of SCl(2) and the (approximate)X(2)B(1), (approximate)A(2)B(2), (approximate)B(2)A(1), (approximate)C(2)A(1), (approximate)D(2)A(2), and (approximate)E (2)B(2) states of SCl(2) (+) at the restricted-spin coupled-cluster single-double plus perturbative triple excitation [RCCSD(T)] level with basis sets of up to the augmented correlation-consistent polarized quintuple-zeta [aug-cc-pV(5+d)Z] quality. Effects of core electron correlation, basis set extension to the complete basis set limit, and relativistic contributions on computed minimum-energy geometrical parameters and/or relative electronic energies were also investigated. RCCSD(T) potential energy functions (PEFs) were calculated for the (approximate)X (1)A(1) state of SCl(2) and the low-lying states of SCl(2)(+) listed above employing the aug-cc-pV(5+d)Z basis set. Anharmonic vibrational wave functions of these neutral and cationic states of SCl(2), and Franck-Condon (FC) factors of the lowest four one-electron allowed neutral photoionizations were computed employing the RCCSD(T)aug-cc-pV(5+d)Z PEFs. Calculated FC factors with allowance for the Duschinsky rotation and anharmonicity were used to simulate the first four photoelectron (PE) bands of SCl(2). The agreement between simulated and observed He I PE spectra reported by Colton et al. [J. Electron Spectrosc. Relat. Phenom. 3, 345 (1974)] and Solouki et al. [Chem. Phys. Lett. 26, 20 (1974)] is excellent. However, our FC spectral simulations indicate that the first observed vibrational component in the first PE band of SCl(2) is a "hot" band arising from the SCl(2)(+)(approximate)X(2)B(1)(0,0,0)<--SCl(2)(approximate)X (1)A(1)(1,0,0) ionization. Consequently, the experimental adiabatic ionization energy of SCl(2) is revised to 9.55+/-0.01 eV, in excellent agreement with results obtained from state-of-the-art ab initio calculations in this work.

Cations, Divalent↗

PubMed2004

Translating knowledge into practice in the "post-genome" era.

The Human Genome Project is "completed", but it is only a beginning in the understanding of genomic structure and function. A "human phenome project" is waiting in the wings. The complexity involving a phenotype can be glimpsed, for example, if one enquires into the relationships between mutant phenylalanine hydroxylase (PAH) genotypes and the clinical disorders called PKU/Hyperphenylalaninemia-so called lessons from PKU genotypes and phenotypes. Since genomes speak biochemistry, not phenotype (said RHA Plasterk), for genomics to penetrate medicine, biochemistry and biology must be allies. The ideal translators and ambassadors of the knowledge that must cross the gap between laboratory and bedside are the clinician scientists; restoration of that attenuated community of colleagues is a necessary step in the implementation of genomic medicine.

Genome↗

PubMed1990

UV light and ionizing radiations cause programmed death of rat chloroleukaemia cells by inducing retropositions of a mobile DNA element (L1Rn).

The long interspersed repetitive DNA, L1 or LINE, is a class of mobile genetic elements which can amplify in the cell genome by retroposition, i.e. by a mechanism similar to that of retroviruses. We have shown before that in rat chloroleukaemia cells, maintained in suspension culture in vitro, this element is spontaneously transcriptionally activated at about half of the maximal population density. About 24 h later an explosive amplification of the L1 element is seen in DNA: about 300,000 copies are inserted into apparently random locations in the cell genome, thus creating an outburst of lethal mutations. Dead cells display morphological features typical to programmed death. The present results show that UV light and ionizing radiation induce rapid, premature activation of the L1Rn element during the fast exponential growth of chloroleukaemia cells, and that also this exogenously induced activation is followed by programmed cell death. Transcriptional activation of the L1Rn element can be very strong after the UV exposure: at least 70-fold. Severe hyperthermia, lethal to the cells, does not lead to L1Rn activation (actually a marked suppression is seen) and the mode of phenomic death is necrosis. Some biological implications of the results are discussed.

Animals↗

PubMed2005

Classification of human population based on HLA gene polymorphism and the concept of Prakriti in Ayurveda.

BACKGROUND: Correlating phenotypes with genotypes remains the major postgenomic challenge. Attempts to correlate phenotype characteristics associated with ethnicity, geographical divisions, or diseases to genotypes have had limited success. This means that current approaches for identifying phenotypes associated with haplotypes may be inadequate. OBJECTIVES: We hypothesize that a human phenome based on Ayurveda could provide an appropriate approach. Specifically, there could be a genetic basis for the three major constitutions (Prakriti) described in Ayurveda. The Prakriti classification is based on differences in physical, physiological, and psychologics characteristics and is independent of racial, ethnic, or geographical considerations. It may provide an appropriate means of classifying phenotypes to be considered collectively for genotyping. DESIGN: As a pilot study to test the hypothesis, we evaluated 76 subjects both for their Prakriti and human leucocyte antigen (HLA) DRB1 types. The genomic DNA was extracted using a standard protocol. Subsequently, HLA DRB1 typing was done by low-resolution polymerase chain reaction sequence specific primers and oligonucleotide probes. RESULTS: We observed a reasonable correlation between HLA type and Prakriti type. The complete absence of the HLA DRB1*02 allele in the Vata type and of HLA DRB1*13 in the Kapha type are significant, with X2 = 4.715 and p < 0.05. HLA DRB1*10 had higher allele frequency in the Kapha type than in the Pitta and Vata types. CONCLUSION: Ayurveda classifies the whole human population in three major constitutions as Vata, Pitta, Kapha and their possible combinations. Their homologous relation to human genetic structure needs to be studied for validation. If validated, our hypothesis would have far reaching implications for pharmacogenomics, modern genetics, human health, and Ayurveda.

Adult↗

PubMed2004

Interactions among factors regulating phenological development and acclimation rate determine low-temperature tolerance in wheat.

BACKGROUND AND AIMS: Exposure to low temperatures (LT) produces innumerable changes in morphological, biochemical and physiological characteristics of plants, with the result that it has been difficult to separate cause and effect adjustments to LT. Phenotypic studies have shown that the LT-induced protective mechanisms in cereals are developmentally regulated and involve an acclimation process that can be stopped, reversed and restarted. The present study was initiated to separate the developmental factors determining duration from those responsible for rate of acclimation, to provide the opportunity for a more in depth analysis of the critical mechanisms that regulate LT tolerance in wheat (Triticum aestivum). METHODS: The non-hardy spring wheat cultivar 'Manitou' and the very cold-hardy winter wheat cultivar 'Norstar' were used to produce reciprocal near-isogenic lines (NILs) in which the vrn-A1 (winter) alleles of 'Norstar' were inserted into the non-hardy 'Manitou' genetic background and the Vrn-A1 (spring) alleles of 'Manitou' were inserted in the hardy 'Norstar' genetic background so that the effects of duration and rate of LT acclimation could be quantified. KEY RESULTS: Comparison of the acclimation curves of the NILs and their parents grown at 2, 6 and 10 degrees C established that the full expression of LT-induced genetic systems was revealed only under genotypically dependent optimum combinations of time and temperature. Both duration and rate of acclimation were found to contribute significantly to the 13.8 degrees C difference in lowest survival temperature between 'Norstar' and 'Manitou'. CONCLUSIONS: Duration of LT acclimation was dependent upon the rate of phenological development, which, in turn, was determined by acclimation temperatures and vernalization requirements. Rate of acclimation was faster for genotypes with the 'Norstar' genetic background but the ability to sustain a high rate of acclimation was dependent upon the length of the vegetative stage. Complex time/temperature relationships and unexplained genetic interactions indicated that detailed functional genomic or phenomic analyses of natural allelic variation will be required to identify the critical genetic components of a highly integrated system, which is regulated by environmentally responsive, complex pathways.

Acclimatization↗

PubMed2004

Back to the biology in systems biology: what can we learn from biomolecular networks?

Genome-scale molecular networks, including protein interaction and gene regulatory networks, have taken centre stage in the investigation of the burgeoning disciplines of systems biology and biocomplexity. What do networks tell us? Some see in networks simply the comprehensive, detailed description of all cellular pathways, others seek in networks simple, higher-order qualities that emerge from the collective action of the individual pathways. This paper discusses networks from an encompassing category of thinking that will hopefully help readers to bridge the gap between these polarised viewpoints. Systems biology so far has emphasised the characterisation of large pathway maps. Now one has to ask: where is the actual biology in 'systems biology'? As structures midway between genome and phenome, and by serving as an 'extended genotype' or an 'elementary phenotype', molecular networks open a new window to the study of evolution and gene function in complex living systems. For the study of evolution, features in network topology offer a novel starting point for addressing the old debate on the relative contributions of natural selection versus intrinsic constraints to a particular trait. To study the function of genes, it is necessary not only to see them in the context of gene networks, but also to reach beyond describing network topology and to embrace the global dynamics of networks that will reveal higher-order, collective behaviour of the interacting genes. This will pave the way to understanding how the complexity of genome-wide molecular networks collapses to produce a robust whole-cell behaviour that manifests as tightly-regulated switching between distinct cell fates - the basis for multicellular life.

Animals↗

PubMed2026

TL-HDMR: a transfer learning framework for advancing equitable causal inference reveals metabolic signatures of stroke across multiple ancestries.

The limited genetic diversity in genome-wide association studies (GWAS) poses a significant challenge to the generalizability and equity of biomedical discoveries. Most causal inferences, particularly from high-dimensional phenomes (e.g. metabolomics), are primarily based on European populations, and their applicability to other ancestries remains uncertain. Traditional multivariable Mendelian randomization (MVMR) methods further struggle in high-dimensional and correlated settings due to collinearity and model instability. To bridge this gap, we present a two-step transfer learning framework for high-dimensional MR (TL-HDMR), designed to enhance causal exposure detection in understudied populations. Our approach leverages the Minimax Concave Penalty for asymptotically unbiased estimation amidst exposure correlations. Crucially, we introduce two novel pre-transfer procedures-HDMR.TSD for sourcing beneficial data and HDMR.PRESSO for filtering pleiotropic instruments-to ensure robust knowledge transfer. Extensive simulations demonstrated TL-HDMR's superior performance in ROC curves and mean absolute error over alternative methods. When applied to identify causal metabolites for stroke across multi-ancestry cohorts (European, East Asian, South Asian, and African), TL-HDMR successfully pinpointed both shared and ethnic-specific causal biomarkers, showcasing its unique capability for equitable causal inference. This work provides a powerful statistical tool that not only addresses critical methodological challenges but also promotes inclusivity and fairness in human health research.

Humans↗