Transcriptome-Wide Association Studies of 81 traits in 79,294 Korean individuals

Transcriptome-Wide Association Studies of 81 traits in 79,294 Korean individuals | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Transcriptome-Wide Association Studies of 81 traits in 79,294 Korean individuals Sungho Won, Min Heo, Yunmi Ji, Ah Ra Do, Heejin Jin, Kyungtaek Park, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6417034/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Genome-wide association studies (GWAS) have linked genetic loci to complex traits, but their interpretation is limited by linkage disequilibrium (LD) and difficulties in identifying causal genes. Transcriptome-wide association studies (TWAS) address these challenges by leveraging predicted gene expression to map genomic risk regions. However, TWAS reference panels are predominantly based on European ancestry, restricting their utility for other populations. Here, we developed a Korean-specific whole blood reference panel for transcriptome imputation and performed TWAS, examining 81 traits in 79,294 Korean individuals. This Korean panel showed improved predictive performance over the GTEx reference panel, identifying 348 significant gene-trait associations (Bonferroni-corrected P < 0.05), including 232 novel associations, compared to 180 associations with GTEx. Furthermore, pathway analysis revealed unique biological processes in the Korean model, offering insights into diseases such as thyroid disease and type 2 diabetes. Drug repurposing identified promising candidates, including clopidogrel for angina and vorinostat for fatty liver. These findings highlight the value of population-specific approaches in genetic research. Biological sciences/Genetics/Genetic association study Biological sciences/Genetics/Gene expression Biological sciences/Genetics/Functional genomics/Gene expression profiling Biological sciences/Genetics/Genomics/Transcriptomics Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Genome-wide association studies (GWAS) have successfully identified numerous genetic loci associated with complex traits. However, interpreting these loci remains challenging due to linkage disequilibrium (LD), which often obscures the causal variants driving these associations. Moreover, GWAS alone typically do not allow to pinpoint the target genes mediating the effects of causal variants on traits, and are constrained by the burden of multiple testing 1 , 2 . To address these interpretability and statistical power issues, transcriptome-wide association studies (TWAS) have emerged as a powerful approach in large-scale association studies 2 – 4 . TWAS integrate genetic data with gene expression levels, using genetically regulated gene expression (GReX) to identify risk genes associated with complex traits and diseases. The standard two-stage TWAS approach involves first constructing gene expression imputation models using a transcriptomic reference panel, where gene expression is treated as the response and genetic variants are used as predictors. Associated variants are called expression quantitative trait loci(eQTLs) their estimated effect sizes are then used as weights in gene-based association tests with GWAS data. By incorporating eQTLs from relevant tissues, TWAS improves interpretability and identifies functionally relevant loci. This gene-centric method consolidates the effects of various regulatory variants into a singular test unit, enhancing the study's power and yielding more interpretable genomic loci associated with traits 2 . Despite its advantages, TWAS reference panels, such as the widely used panel of the Genotype-Tissue Expression (GTEx) project, are predominantly derived from individuals of European ancestry. This lack of diversity limits the applicability of TWAS to other populations 5 , 6 , as population-specific differences in allele frequencies, LD patterns, and haplotypes can reduce the accuracy of genetic signal detection, leading to false negatives in non-European populations 7 , 8 . This bias has hindered the comprehensive understanding of complex traits across diverse populations and ancestries. Recent studies underscore the importance of incorporating ancestry-specific models to improve accuracy in genetic association studies across global populations 7 , 9 . In this study, we developed a reference panel specifically for the Korean populations and compared its performance with the GTEx panel, which is based predominantly on European individuals. Using these panels, we conducted a large-scale TWAS on 81 traits in 79,294 Korean individuals. Our analysis identified 348 significant gene-trait associations (Bonferroni-corrected P < 0.05), including 232 novel associations using the Korean panel, compared to 180 associations using the GTEx panel. Furthermore, using these significant gene-trait associations, pathway enrichment analysis revealed distinct biological processes that were enriched in the Korean-based model, such as thyroid disease, type 2 diabetes, and hyperlipidaemia that were distinct from those detected using the GTEx panel. Additionally, computational drug repurposing prioritized promising therapeutic candidates, such as clopidogrel for angina and vorinostat for fatty liver disease, highlighting the translational potential of these findings. These results underscore the importance of population-specific approaches in uncovering novel genetic associations and provide critical insights into the genetic architecture of complex traits in underrepresented populations. Our study emphasizes the necessity of diverse reference panels to advance precision medicine and improve global health equity. Results Overview of the Study In this study, we performed GReX modeling by integrating transcriptomic data extracted from whole blood tissue in two Korean cohorts (Asan and Chosun cohorts). We then compared the results from Korean-based GReX model with the GTEx-based GReX model, which was primarily derived from European individuals. Next, we applied both models to genotype data from the Korean Genome and Epidemiology Study (KoGES) 10 and and the Gene-environmental interaction and phenotype (GENIE) 11 cohorts (total N = 79,294) and conducted TWAS for 81 phenotypes. To evaluate the novelty and biological relevance of the TWAS-identified gene-trait associations, we utilized the Harmonizome 12 database and conducted pathway enrichment analyses 13 . Finally, we employed the CMap linked user environment (CLUE) 14 for computational drug repurposing to identify bioactive small molecules capable of reversing the expression profiles of clinically relevant trait-associated genes 15 (Fig. 1 ). Comparison of GReX Models We focused on genes meeting two criteria: an average Pearson correlation coefficient between predicted and observed gene expression > 0.1 and P < 0.05 in the association test between predicted and observed gene expression based on Stouffer’s method (see Methods). The GTEx-based model included 7,252 genes, whereas the Korean-based model comprised 4,699 genes (Fig. 2 A). When evaluating performance on the respective training datasets, the GTEx-based model outperformed the Korean-based model ( \(\:{R}^{2}=\) 0.11 vs \(\:{R}^{2}=\) 0.07). However, when evaluated on the Korean data, the GTEx-based model underperformed compared to the Korean-based model ( \(\:{R}^{2}=\) 0.03 vs. \(\:{R}^{2}=\) 0.07) (Fig. 2 B). This trend held for intersecting genes, with the Korean-based model showing better performance on Korean data ( \(\:{R}^{2}=\) 0.08 vs \(\:{R}^{2}=\) 0.06) (Fig. 2 B). To further evaluate the models, we calculated the cis-heritability of the genes in both model on the Korean data. The mean of cis-heritability for genes in the Korean-based model was higher compared to the GTEx-based model ( \(\:{h}^{2}=\) 0.09 vs \(\:{h}^{2}=\) 0.05). Notably, the genes common to both models exhibited an even higher mean cis-heritability ( \(\:{h}^{2}=\) 0.11) (Fig. 2 C, Table S1 ). In addition, allele frequencies of SNPs differed by population: SNPs in the Korean-based model had lower frequencies in East Asians, whereas those in the GTEx-based model had lower frequencies in Europeans 16 , 17 (Figure S1 ). An eQTL analysis revealed that 94% of genes in the Korean-based model (4,434/4,699) were significantly associated with SNPs (FDR-adjusted P < 0.05), compared to 49% (3,553/7,252) in the GTEx-based model. Similarly, the Korean-based model had a higher proportion of significant eQTL-gene pairs (54.8% vs 19.3%) (Table S2). Associations of Predicted Gene Expression with 81 Traits Of the 81,153 individuals in KoGES + GENIE cohorts, 79,294 with no missing data for age, sex, BMI, and smoking status were considered for TWAS for 81 traits. We assessed both the Korean-based and GTEx-based models, finding that their correlations of predicted gene expression obtained from intersecting genes was relatively high ( \(\:{R}^{2}=\) 0.40), with most correlations being positive (Figure S2). TWAS was then conducted for genes with \(\:{R}^{2}\) (see Methods) greater than 0.05, yielding 1,104 genes from the GTEx-based model (out of 7,252 total) and 1,630 genes from the Korean-based model (out of 4,699 total). Using a Bonferroni-adjusted P threshold, the Korean-based model identified 348 significant gene-trait associations (181 genes across 37 traits), whereas the GTEx-based model identified 260 (139 genes across 34 traits). Among these, 138 associations (72 genes across 31 traits) were significant in both models (Fig. 3 , Table S7). Notably, the Korean-based model identified more significant genes than the GTEx-based model in most traits (33 out of 39 traits; Figure S3). Detailed trait information is provided in Table S3. We also examined the significance and directionality of TWAS \(\:Z\) -scores for gene-trait associations common to both models. Most significant gene-trait associations were positive and unique to the Korean-based model (not significant in GTEx-based model), followed by negative associations unique to the Korean-based model. Additionally, \(\:Z\) -scores for overlapping gene-trait associations showed consistent directionality (Figure S4), with a correlation of 0.4229 across two models (Figure S5). Identification of Novel Gene-Trait Associations Using the Harmonizome database 12 , we found that 137 gene-trait associations unique to the Korean-based model were novel (44 strong, 55 weak, 38 no). In contrast, the GTEx-based model identified 85 novel associations (22 strong, 17 weak, 46 no). Among shared associations, 95 were novel (32 strong, 23 weak, and 40 no) (Table S5, Table S6). Pathway Enrichment Analysis For diseases associated with at least three genes, we performed a gene ontology (GO) analysis. GO analysis identified 369 enriched biological processes using the Korean-based model, and 294 using the GTEx-based model (FDR P < 0.1). Clustering analysis with REVIGO 18 highlighted distinct biological patterns. Both models identified pathways for type 2 diabetes and hyperlipidemia, but with distinct enriched terms (Fig. 4 , Table S8). For type 2 diabetes, the Korean-based model highlighted terms such as ‘O-glycan processing’ (34.6%), ‘female gonad development’ (18.1%), ‘negative regulation of muscle hypertrophy’ (11.5%), and ‘natural killer cell-mediated cytotoxicity directed against tumor cell targets’ (11.5%). In contrast, the GTEx-based model uncovered distinct terms, including ‘RNA splicing’ (27.6%), ‘negative regulation of antibody-dependent cellular cytotoxicity’ (19.1%), ‘response to fluoride’ (18.1%), and ‘specification of segmental identity antennal segment’ (14.1%). Notably, no overlapping clusters were identified between the two models for type 2 diabetes (Fig. 4 C). Similarly, both models produced significant results for hyperlipidemia, but again with non-overlapping enriched terms. The Korean-based model identified terms such as ‘endonucleolytic cleavage to generate mutated 5’-end of SSU-rRNA’ (26.0%), ‘female meiosis chromosome segregation’ (17.7%), ‘nucleobase-containing compound transport’ (13.4%), and ‘regulation of response to biotic stimulus’ (12.8%). In contrast, the GTEx-based model identified different terms such as ‘regulation of biosynthetic process of antibacterial peptides active against Gram-positive bacteria’ (37.2%), ‘synaptic vesicle targeting’ (22.0%), ‘enzyme-directed rRNA pseudouridine synthesis’ (19.6%), and ‘inflorescence meristem growth’ (12.8%) (Fig. 4 D). In addition, significant terms for thyroid disease were identified only by the Korean-based model, whereas hypertension terms were exclusive to the GTEx-based model. For thyroid disease, the top enriched GO terms in the Korean-based model included ‘auxin biosynthetic process’ (40.1%), ‘negative regulation of DNA damage checkpoint’ (18.7%), ‘inflorescence meristem growth’ (10.7%), and ‘histidine transport’ (10.1%) (Fig. 4 B). In contrast, hypertension-related terms in the GTEx-based model involved terms such as somatic diversification of immunoglobulins’ (60.1%), ‘auxin export across the plasma membrane’ (14.0%), ‘negative regulation of atrichoblast fate specification’ (7.1%), and ‘positive regulation of post-transcriptional gene silencing’ (10.1%) (Fig. 4 B). Detailed pathway enrichment results, including GO 19 , KEGG 20 , Reactome 21 , WikiPathways 22 analyses are provided in Table S8. Computational Drug Repurposing Analysis We prioritized drug candidates with therapeutic potential using Connectivity Map (CMap) database derived from the L1000 assay 14 . Using the Korean-based model, we identified significant drug-disease associations. Clopidogrel was associated with angina, vorinostat with fatty liver disease, fulvestrant with periodontal disease, rucaparib with prostate cancer, and pioglitazone with gastritis. Similarly, the GTEx-based model revealed additional significant drug-disease associations. Etoposide was linked to arthritis, enalapril to breast cancer, minoxidil to heart failure, cefotaxime to chronic pulmonary disease, azacitidine to colorectal cancer, simvastatin to lung cancer, and acyclovir to prostate cancer. Only the top-ranked (top 1) drug for each disease was subjected to a literature review (Table 1 ) and the comprehensive list of identified drug-disease pairs is provided in Table S9. Table 1 Top 1 drug repurposing results. Drug enrichment analysis of TWAS results indicates biologically relevant drugs. We imported drug target genes from CLUE and tested whether these gene sets were enriched by TWAS hits. We only displayed Top 1 drug repurposing results per phenotype that satisfied a connectivity score (tau) <= -90 according to each model. All significant results can be found in Table S9. Model Phenotype Drug name Mechanism of Action Korean_Tau GTEx_Tau Korean angina clopidogrel purinergic receptor antagonist -91.02 * -88.45 asthma panobinostat HDAC inhibitor -98.10 * 0.00 prostate hyperplasia prednicarbate phospholipase activator -96.69 * -9.57 chronic pulmonary disease bifonazole sterol demethylase inhibitor -95.38 * -1.93 colorectal cancer digoxin ATPase inhibitor -94.16 * -70.19 type 2 diabetes flunisolide cytochrome P450 inhibitor -95.33 * 0.00 fatty liver vorinostat HDAC inhibitor -98.29 * -2.68 fracture tivozanib VEGFR inhibitor -94.623 * 11.27 chronic gastritis etodolac cyclooxygenase inhibitor -94.41 * -4.86 gallstones pitavastatin HMGCR inhibitor -94.51 * -10.39 gastric cancer triamcinolone glucocorticoid receptor agonist -90.62 * -80.08 liver cancer deferiprone chelating agent -94.70 * -68.52 chronic hepatitis C panobinostat HDAC inhibitor -95.17 * 9.74 hypertension selumetinib MEK inhibitor -94.54 * -0.14 hyperlipidemia clotrimazole cytochrome P450 inhibitor|imidazoline receptor ligand -90.67 * -74.17 acute liver disease propofol benzodiazepine receptor agonist -94.72 * -11.05 myocardial infarction oxiconazole bacterial cell wall synthesis inhibitor -93.57 * 8.25 periodontal disease fulvestrant estrogen receptor antagonist -95.77 * -96.23 * prostate cancer rucaparib PARP inhibitor -92.87 * -1.76 gastritis, gastric/duodenal ulcer pioglitazone insulin sensitizer|PPAR receptor agonist -91.22 * -2.08 tuberculosis noscapine bradykinin receptor antagonist|tubulin polymerization inhibitor -90.30 * -31.46 thyroid cancer digoxin ATPase inhibitor -97.74 * 0.00 transient ischemic attack (TIA) fulvestrant estrogen receptor antagonist -95.02 * -18.94 cystitis homatropine acetylcholine receptor antagonist -90.02 * 18.86 valvular disease albendazole tubulin polymerization inhibitor -93.80 * -0.18 GTEx angina docetaxel tubulin polymerization inhibitor 95.21 -90.37 * arrhythmia vinorelbine tubulin polymerization inhibitor -17.90 -95.88 * arthritis etoposide topoisomerase inhibitor -32.14 -94.08 * asthma vinorelbine tubulin polymerization inhibitor -73.19 -94.49 * breast cancer enalapril angiotensin converting enzyme inhibitor -10.27 -90.38 * coronary artery disease propranolol adrenergic receptor antagonist -6.20 -92.49 * cataract liothyronine thyroid hormone stimulant -73.42 -96.70 * heart failure minoxidil KATP activator|Kir6 channel (KATP) activator|vasodilator -13.03 -99.15 * liver cirrhosis fluvoxamine selective serotonin reuptake inhibitor (SSRI) 0.00 -98.22 * chronic pulmonary disease cefotaxime bacterial cell wall synthesis inhibitor -4.48 -91.80 * colorectal cancer azacitidine DNA methyltransferase inhibitor 0.00 -93.69 * cerebrovascular disease olanzapine dopamine receptor antagonist|serotonin receptor antagonist 5.81 -92.42 * type 2 diabetes deferiprone chelating agent -45.94 -99.12 * fatty liver sirolimus mTOR inhibitor -2.61 -91.95 * fracture metformin insulin sensitizer 0.00 -96.25 * chronic gastritis vinorelbine tubulin polymerization inhibitor -74.89 -93.15 * gallstones sulpiride dopamine receptor antagonist -44.98 -95.08 * gout vinorelbine tubulin polymerization inhibitor -69.65 -94.90 * chronic hepatitis B vidarabine antiviral -4.23 -97.11 * hypertension vinorelbine tubulin polymerization inhibitor -85.45 -95.87 * kidney disease lypressin vasopressin receptor agonist 0.00 -94.44 * lung cancer simvastatin HMGCR inhibitor 0.00 -95.10 * hyperlipidemia docetaxel tubulin polymerization inhibitor -39.99 -90.12 * acute liver disease diclofenamide carbonic anhydrase inhibitor -0.25 -93.92 * myocardial infarction rifabutin protein synthesis inhibitor -8.09 -91.87 * periodontal disease eugenol androgen receptor antagonist -13.69 -97.65 * prostate cancer acyclovir DNA polymerase inhibitor 0.00 -92.25 * peripheral vascular disease vinorelbine tubulin polymerization inhibitor -6.87 -90.26 * smoking status docetaxel tubulin polymerization inhibitor -87.36 -90.30 * tuberculosis deferiprone chelating agent -44.71 -92.63 * thyroid disease albendazole tubulin polymerization inhibitor 15.74 -90.60 * thyroid cancer vincristine tubulin polymerization inhibitor -13.00 -91.21 * transient ischemic attack oxybutynin acetylcholine receptor antagonist 15.04 -98.17 * cystitis vecuronium acetylcholine receptor antagonist -4.97 -99.89 * valvular disease benperidol dopamine receptor antagonist -14.98 -92.25 * Korean_Tau: Connectivity score of durg repurposing using genes with significant association with each phenotype when TWAS was performed based on Korean-based model GTEx_Tau: Connectivity score of durg repurposing using genes with significant association with each phenotype when TWAS was performed based on GTEx-based model Discussion In this study, we conducted TWAS for 81 traits in 79,294 Korean individuals by applying both a GReX model specifically trained for Koreans based on whole blood tissue and an existing GTEx-based model. We identified 470 significant gene-trait associations (Bonferroni P < 0.05), of which 317 were novel. Pathway enrichment analysis revealed both shared and unique biological processes in the Korean-based and GTEx-based models, with the former highlighting thyroid disease and the latter emphasizing hypertension. Additionally, our computational drug repurposing analysis uncovered several promising therapeutic candidates, including clopidogrel for angina and vorinostat for fatty liver disease. Our TWAS analysis utilizing a Korean-specific GReX model identified several novel gene-trait associations, highlighting potential mechanistic links based on existing literature. ANXA3, a regulator of cellular stress responses, may contribute to kidney dysfunction, potentially influencing albumin homeostasis indirectly 23 , 24 . CUL7, a key regulator of skeletal growth, was associated with height, aligning with its known role in 3M syndrome 25 , 26 . GPSM3, an immune signaling modulator, may affect HbA1c levels through inflammation-driven insulin resistance 27 , 28 . ITIH4, an inflammation-related plasma glycoprotein, exhibited a negative correlation with inflammatory processes and coagulation, both of which are known to impact hematocrit and erythropoiesis 29 – 31 . Lastly, MT2A, a metal-binding protein, may modulate HDL-C levels by enhancing antioxidant capacity and protecting against lipid peroxidation 32 – 34 . Subsequent pathway enrichment analysis provided further evidence of disease-related mechanisms, suggesting that certain genes within these clusters may serve as causal factors in relevant biological processes or pathways. For instance, recent evidence for thyroid disease points to possible antioxidant effects of the plant hormone auxin, which can protect thyroid membrane lipids from oxidative damage in animal models. Although auxin biosynthesis in plants is well-established, the potential connection between plant hormone and human endocrine remains underexplored. Plant hormones and their derivatives exhibit therapeutic potential, but may disrupt thyroid function if not properly managed, emphasizing the need for careful monitoring and ongoing research. Moreover, the negative regulation of DNA damage checkpoints is increasingly recognized in thyroid disease, particularly in thyroid tumors and endocrine regulation 35 – 37 . Disruptions in histidine transport have also been linked to thyroid hormone imbalances, suggesting an additional avenue for endocrine dysfunction 38 , 39 . For hypertension, the positive regulation of post-transcriptional gene silencing has identified microRNAs as crucial mediators in blood pressure regulation and cardiovascular health. MicroRNAs, which are involved in post-transcriptional gene silencing by targeting specific mRNAs for degradation or translational repression, play a significant role in hypertension, as observed in studies highlighting their regulatory impact on these biological pathways 40 , 41 . For type 2 diabetes, alterations in O-glycan processing, particularly in O-GlcNAc modifications, influence insulin signaling and secretion, playing a critical role in the development of type 2 diabetes 42 , 43 . Additionally, a connection between female gonad development and type 2 diabetes has been observed, as hormonal pathways in conditions like polycystic ovary syndrome (PCOS) illustrate how androgen excess can affect both reproductive and metabolic health 44 – 46 . The negative regulation of muscle hypertrophy, exemplified by elevated levels of myostatin, may exacerbate insulin resistance and contribute to metabolic dysfunction in type 2 diabetes 47 . Furthermore, disruptions in RNA splicing processes are closely associated with insulin resistance, beta-cell dysfunction, and the onset of type 2 diabetes, offering potential targets for early diagnosis and therapeutic interventions 48 . There is also evidence that fluoride exposure might impact glucose metabolism and insulin sensitivity, with effects dependent on the dose, duration, and presence of existing metabolic conditions 49 , 50 . For hyperlipidemia, there is evidence that lipid imbalance can alter the immune regulation by disrupting T cell function and promoting chronic inflammation, further connecting lipid metabolism and immune homeostasis 51 , 52 . The results of our computational drug repurposing analysis identified several repurposed drugs with potential therapeutic applications. Using the Korean-based model, clopidogrel, a purinergic receptor antagonist, was found to reduce thrombotic events in angina by inhibiting the P2Y12 receptor and decreasing platelet aggregation 53 , 54 . Vorinostat, an HDAC inhibitor, demonstrated potential efficacy in the treatment of fatty liver disease through gene expression modulation 55 . Fulvestrant, an estrogen receptor antagonist, may influence fibroblast proliferation and inflammation, key contributors to periodontal disease progression 56 – 58 . Rucaparib, a PARP inhibitor, emerged as a promising therapy for metastatic castration-resistant prostate cancer, particularly in BRCA-mutated cases 59 . Pioglitazone, while lacking direct evidence for treating gastritis, may confer anti-inflammatory and antioxidant properties that alleviate gastric inflammation and mucosal damage 60 . In contrast, the GTEx-based model identified etoposide, a topoisomerase inhibitor, as potentially reducing inflammation in arthritis by modulating immune function 61 . Enalapril, an angiotensin-converting enzyme inhibitor, showed chemopreventive properties in breast cancer by affecting angiogenesis and inflammation 62 . Minoxidil, a KATP channel activator and vasodilator, demonstrated cardioprotective effects in heart failure 63 , 64 . Cefotaxime, a bacterial cell wall synthesis inhibitor, may be beneficial in chronic pulmonary disease by targeting bacterial infections 65 while azacitidine, a DNA methyltransferase inhibitor, reactivates tumor suppressor genes and enhances immune responses in colorectal cancer 66 , 67 . Simvastatin, an HMGCR inhibitor, demonstrated anti-cancer properties in lung cancer by reducing cell proliferation and metastasis 68 , 69 , and acyclovir has also shown promise in treating prostate cancer via inhibiting DNA replication 70 , 71 . Beyond known treatments like clopidogrel for angina and vorinostat for fatty liver disease, our analysis suggested novel mechanisms of existing drugs. For instance, HDAC inhibition, predominantly studied in oncology, may be effective in non-cancerous conditions such as asthma 72 . Similarly, glucocorticoid receptors modulation, typically linked to inflammation, could be beneficial in gastric cancer, despite poorer prognoses associated with overexpression in certain tissues 73 , 74 . Chelation therapy, particularly targeting metals like zinc and copper, demonstrates potential for managing type 2 diabetes by reducing oxidative stress and improving insulin sensitivity 75 , 76 , and it could also lower liver cancer risk 77 , 78 . Targeting tubulin polymerization has emerged as a strategy against drug-resistant tuberculosis and some cancers, including thyroid cancer, by disrupting cellular structure and division 79 , 80 whereas inhibition of ATPase pathways - crucial in hypoxia signaling – may be applicable to thyroid cancer treatment 81 . Acetylcholine receptor antagonists could regulate pain and inflammation in cystitis models, suggesting its therapeutic potential in bladder conditions 82 , 83 . Furthermore, mTOR inhibition, particularly mTORC1, shows potential in treating fatty liver disease by reducing lipid accumulation and enhancing fatty acid metabolism 84 , 85 . Although this study focused on the top candidate drug for each phenotype, future research should explore combination therapies and alternative drugs with slightly lower connectivity scores to uncover synergistic benefits or novel therapeutic strategies. Further experimental validation will be necessary to confirm the therapeutic potential of these repurposed drugs in various disease contexts. Despite these promising findings, our study has certain limitations. First, our reliance on whole blood tissue as the sole source for the GReX model may skew results for traits less relevant to this tissue, such as thyroid cancer. Including multi-tissue or cell-type specific data would address this limitation 2 . Second, we constructed our Korean GReX model using imputed SNP chip data rather than whole-genome sequencing. Although we applied stringent quality control measures (imputation info score > 0.8), imputation-related errors and biases in effect size estimates remain potential challenges 86 . Finally, the lack of extensive replication analyses for novel signals limits the generalizability of our findings. Further replication involving larger East Asian cohorts are necessary to validate these associations. In conclusion, our large-scale TWAS on a Korean population offers important insights into the genetic architecture of complex traits, uncovering multiple novel gene-trait associations with implications for disease prevention, diagnosis, and treatment. The pathway-level analyses underscore the value of ancestry-specific models, which can reveal previously unrecognized mechanisms and deepen our understanding of how genetic factors contribute to complex diseases. By integrating computational approaches, we have also identified opportunities to repurpose existing drugs, potentially expediting the translation of genetic discoveries into clinical applications. Ultimately, by focusing on underrepresented population, our research underscores the importance of incorporating diverse genetic data to enhance the accuracy, relevance, and inclusivity of genetic studies. This study not only advances our understanding of population-specific genetic architectures but also lays the foundation for globally impactful discoveries that can inform precision medicine and improve health outcomes for diverse populations. Methods GReX Training Dataset Information We combined two datasets for Korean GReX model training. The first dataset included genotypes and whole blood RNA sequencing data from 436 Korean subjects with mild asthma symptoms like dyspnea, wheezing, or cough for over three months, and airway hyperresponsiveness but without severe lung damage, bronchiectasis, or a history of lung resection 87 . The second dataset was drawn from participants in the Gwangju Alzheimer’s & Related Dementia (GARD) cohort registry at Chosun University in Gwangju, Korea, and included genotypes and RNA sequencing data from 300 individuals with mild cognitive impairment 88 . Genotype and RNA Quality Control in GReX Training Dataset Asan and Chosun study participants were genotyped using the Korea Biobank Array, which was designed by the Korea National Institute of Health (KNIH) 89 . Both datasets were preprocessed as follows. A K-medoid clustering algorithm was employed to identify the genotypes, minimizing batch effects 90 . Quality control procedures were executed using PLINK and ONETOOL 91 , with exclusion criteria set for samples with sex discrepancies or call rates below 90%. SNPs were filtered based on a call rate threshold of less than 95% or deviation from the Hardy–Weinberg equilibrium (HWE) test ( P < 1 × 10 − 6 ). Imputation was conducted with IMPUTE2 92 using the NARD V1 imputation server 93 . Pre-phasing was performed with Eagle v2.4.1 94 , and imputation was done with Minimac4 95 , both using default settings. Post-imputation quality control was used to exclude imputed SNPs with \(\:{R}^{2}\) 0.05, HWE P < 1 × 10 −6 , and minor allele frequencies (MAF) <0.01. After QC, SNP coordinates were updated to the hg38 assembly using the R package “rtracklayer” (version 1.46.0). Then, the two data sets were merged and SNPs were excluded with missing genotype rates > 0.05, HWE P < 1 × 10 −6 , and MAF < 0.01. Finally, 4,535,155 SNPs for 674 samples were used for GReX modeling. RNA sequencing data were also available for these Asan and Chosun study participants, with detailed descriptions of RNA library preparation and sequencing provided in the previous publications 87 , 88 . Briefly, transcriptomic QC involved trimming low-quality reads and adapter sequences with Trimmomatic v0.39 96 . The filtered reads were aligned to the GRCh38/hg38 human reference genome using STAR v2.7.8a 97 , and gene-level expression was quantified using RNA-SeQC v2.4.2 98 with GENCODE Release 26 annotation (gencode.v26.GRCh38.genes.gtf) 99 . Genes with a Transcripts per Million (TPM) value < 0.1 in more than 80% of the participants were excluded. The resulting gene expression profiles were normalized by TMM 100 , followed by inverse normal transformation. To correct for batch effects and experimental variability, PEER 101 factor analysis was conducted. Finally, normalized gene expression values were adjusted for sex, the top 3 genotype principal components(PCs), and the top 60 PEER factors, as recommended by the GTEx Consortium 102 . Genotype Quality Control for TWAS Study Samples We conducted TWAS on the KoGES and GENIE cohorts, both of which were genotyped using the Korea Biobank Array designed by the Korea National Institute of Health (KNIH) 89 . Variant calling, quality control (QC), and imputation procedures followed the same protocol used for the GReX training dataset. After merging both cohorts, 3,504,297 SNPs across 81,153 subjects remained for analysis. Building GReX Prediction Models We constructed the predictive models for GReX by integrating genotype and transcriptome data (Asan + Chosun) and focusing on cis-region SNPs (± 1 Mb of a gene’s start or end position) for downstream TWAS analysis. Model development was conducted using the PredictDB pipeline, which automates data preprocessing, training, and integration of the elastic net regression results into database files used as weights for predicting gene expression 3 This pipeline utilizes the R package “glmnet” (version 4.1.2), designed to model gene expression predictions based on cis-region SNPs 103 . The elastic net regularization is controlled by the shrinkage parameter alpha (α), where α = 0 corresponds to ridge regression and α = 1 corresponds to lasso regression. In our model, α was set to 0.5. Sex, age, 3 genotype-based PCs, and 60 PEER factors were included as covariates to account for possible confounding. To evaluate the predictive accuracy, we employed a nested cross-validation. Our dataset was partitioned into five equally sized folds. For each fold, we applied a 10-fold cross-validated elastic net model to 80% of the data, simultaneously tuning the lambda parameter. We then assessed model performance predicted on the remaining 20% of the data by comparing predicted and observed gene expression to compute the coefficient of determination (𝑅 2 ). We repeated this process for all five folds, ultimately deriving the average 𝑅 2 , the Pearson correlation coefficient, and a combined Z score via Stouffer’s method as metrics of overall model performance across the five folds. Cis-eQTL analysis We performed a cis-eQTL analysis in the previously described GReX training dataset. Cis-region SNPs were defined as those residing within ± 1 Mb of a gene’s start or end position, according to the GENCODE Release 26 (hg38). To identify cis-eQTLs, we employed the Matrix eQTL R package (version 3.2) 104 with default settings, which applies a linear regression framework to test SNP-gene expression associations. In this analysis, gene expression values were adjusted for sex, age, 3 genotype-based PCs, and 60 PEER factors to account for confounding factors. Cis-heritability calculation of gene expression We constructed a genetic relationship matrix for each gene using SNPs with MAF > 0.05, located within ± 1 Mb of the gene and meeting HWE ( P > 1 × 10 − 6 ). From these genetic relationship matrixes, we applied a variance-component model to estimate narrow-sense heritability (ℎ 2 ) of gene expression for each gene. This analysis was performed separately for both the GTEx-based model and the Korean (Asan + Chosun) model. We used the restricted maximum likelihood (REML) approach as implemented in the GCTA 105 software (GCTA-GREML). For the heritability calculations, gene expression values were adjusted for sex, age, 3 genotype-based PCs, and 60 PEER factors. Transcriptome-Wide Association Study (TWAS) of 81 Traits For association analysis, we retained only those genes with a cross validated \(\:{R}^{2}>0.05\) in each model, yielding 1,146 genes for GTEx and 1,678 genes for Korean-based model. Quantitative traits were inverse rank-normalized using the R package “RNOmni” (version 1.0.0), except for anthropometric and lifestyle traits. We performed TWAS on 30 quantitative traits using linear regression and on 51 binary traits using Firth’s logistic regression through the R package “logistf” (version 1.24.1) to mitigate type I error due to case-control imbalance. Multiple testing was addressed by applying separate Bonferroni-corrected significance thresholds for each model: P < 4.36×10 −5 (0.05/1,146) for the GTEx-based model and P < 2.97×10 − 5 (0.05/1,678) for the Korean-based model. All regression models were adjusted for age, sex, body mass index (BMI), smoking status, and the top 10 PCs. Identification of Novel Gene-Trait Associations and Replication We used Harmonizome 12 to assess whether any of the gene-trait associations identified in our TWAS results had been previously reported. Using Harmonizome, we categorized gene-trait associations into four categories: Complete, Strong, Weak, and No (Table S6). This categorization enabled us to systematically identify and propose novel gene-trait associations. Pathway Enrichment Analysis We performed pathway enrichment analysis to evaluate the biological relevance of genes identified through TWAS for each disease (FDR P < 0.1). As the primary resource for enrichment, we first considered the GO 19 gene set collection from the C5 category in MSigDB 13 . Only traits associated with at least three genes in gene-trait association tests were included in the analysis. We used a one-sided Fisher’s exact test to determine the statistical significance of GO terms, with P adjusted via Benjamini-Hochberg method (FDR P < 0.1). Next, we employed REVIGO 18 to reduce redundancy among enriched GO terms, and the refined results were visualized using CirGO 106 for a more intuitive and comprehensive presentation. REVIGO applies a clustering algorithm based on semantic similarity drawing upon pre-computed information content to streamline the interpretation and visualization of GO analysis outcomes. This approach selects a representative subset of terms, facilitating clearer insights into the underlying biological processes. These enrichment analyses were repeated for KEGG 20 , Reactome 21 , and WikiPathways 22 , focusing on canonical pathways from the C2 category in MSigDB 13 . Computational Drug Repurposing Analysis We conducted a computational drug repurposing analysis using the CMap approach 14 , a web-based tool for drug repositioning that matches input gene signatures (both up- and down-regulated) with drug-induced gene expression signatures derived from in vitro experiments ( https://clue.io/ ). For each trait, we selected the top 50 up-regulated and bottom 50 down-regulated genes identified from the TWAS for drug repurposing analysis. Connectivity scores were calculated using Kolmogorov-Smirnov statistics, measuring the association between the TWAS-derived gene expression profile for a given phenotype and the corresponding drug-induced gene expression profile in the CMap database. To ensure clinical relevance, we restricted our analyses to FDA-approved drugs using DrugBank (version 5.1.12, https://go.drugbank.com ) 107 , and obtained their mechanisms of action from the CLUE Drug Repurposing Hub 15 . Ultimately, we identified compounds whose gene expression profiles opposed the disease-risk genes detected through TWAS. Ethical approval and informed consent This study was approved by Institutional Review Board of Seoul National University (E2410/001–004). All participants or their legal guardians provided written informed consent at the time of cohort enrollment. The study using data from Asan Medical Center was approved by the Institutional Review Board (IRB) of Asan Medical Center (2019 − 0376), and the study using data from Chosun University Hospital, Korea (CHOSUN 2013–12–018–070). For further details, please refer to previous studies 87 , 88 . Declarations Data availability Individual-level phenotypic and genetic data for KOGES (https://is.kdca.go.kr/) are available upon application. PrediXcan Elastic-Net base models trained on GTEx V8 can be accessed from the PredictDB Data Repository (https://predictdb.org/). The weights for the Korean GReX model and summary statistics for all TWAS analyses are available upon request. Code availability All code and computational tools used in this study are publicly available. Authors’ contributions MH, SK, and SW conceptualized and designed the study. MH, YJ, AD, TK, JG, KL, SK and SW contributed to data acquisition and analysis. MH, HJ, KP, NK, JA, SK, SW carried out the statistical analysis and interpretation. MH, SK, and SW wrote and revised the manuscript. All authors have provided critical review and edit for the manuscript. Acknowledgements This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government(MSIT)(RS-2024-00346850) and by the National Research Foundation (NRF) grant (NRF-2021R1A5A1033157) funded by the Korean government (MSIT). This work was conducted with bioresources from National Biobank of Korea, the Korea Disease Control and Prevention Agency, Republic of Korea (NBK-2020-101). This work was supported by the national supercomputing center with supercomputing resources including technical support (KSC-2023-CHA-0033), and large data transfer was supported by KREONET, which is managed and operated by KISTI. Min Heo was supported by the Hyundai Motor Chung Mong-Koo Foundation. References Gallagher, M. D. & Chen-Plotkin, A. S. The Post-GWAS Era: From Association to Function. Am J Hum Genet 102 , 717–730 (2018). Wainberg, M. et al. Opportunities and challenges for transcriptome-wide association studies. Nat Genet 51 , 592–599 (2019). Gamazon, E. R. et al. A gene-based association method for mapping traits using reference transcriptome data. Nat Genet 47 , 1091–1098 (2015). Gusev, A. et al. Integrative approaches for large-scale transcriptome-wide association studies. Nat Genet 48 , 245–252 (2016). Ishigaki, K. et al. Large-scale genome-wide association study in a Japanese population identifies novel susceptibility loci across different diseases. Nat Genet 52 , 669–679 (2020). Mai, J., Lu, M., Gao, Q., Zeng, J. & Xiao, J. Transcriptome-wide association studies: recent advances in methods, applications and available databases. Commun Biol 6 , 1–10 (2023). Wojcik, G. L. et al. Genetic analyses of diverse populations improves discovery for complex traits. Nature 570 , 514–518 (2019). Traylor, M. & Lewis, C. M. Genetic discovery in multi-ethnic populations. European Journal of Human Genetics 24 , 1097–1098 (2016). Mogil, L. S. et al. Genetic architecture of gene expression traits across diverse populations. PLOS Genetics 14 , e1007586 (2018). Kim, Y., Han, B.-G., & the KoGES group. Cohort Profile: The Korean Genome and Epidemiology Study (KoGES) Consortium. International Journal of Epidemiology 46 , 1350 (2017). Lee, C. et al. Health and Prevention Enhancement (H-PEACE): a retrospective, population-based cohort study conducted at the Seoul National University Hospital Gangnam Center, Korea. BMJ Open 8 , e019327 (2018). Rouillard, A. D. et al. The harmonizome: a collection of processed datasets gathered to serve and mine knowledge about genes and proteins. Database 2016 , baw100 (2016). Liberzon, A. et al. The Molecular Signatures Database Hallmark Gene Set Collection. Cell Systems 1 , 417–425 (2015). Subramanian, A. et al. A Next Generation Connectivity Map: L1000 Platform and the First 1,000,000 Profiles. Cell 171 , 1437-1452.e17 (2017). Corsello, S. M. et al. The Drug Repurposing Hub: a next-generation drug library and information resource. Nat Med 23 , 405–408 (2017). Karczewski, K. J. et al. The mutational constraint spectrum quantified from variation in 141,456 humans. Nature 581 , 434–443 (2020). Karczewski, K. J. et al. The ExAC browser: displaying reference data information from over 60 000 exomes. Nucleic Acids Research 45 , D840–D845 (2017). Supek, F., Bošnjak, M., Škunca, N. & Šmuc, T. REVIGO Summarizes and Visualizes Long Lists of Gene Ontology Terms. PLOS ONE 6 , e21800 (2011). Ashburner, M. et al. Gene Ontology: tool for the unification of biology. Nat Genet 25 , 25–29 (2000). Kanehisa, M. & Goto, S. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Research 28 , 27–30 (2000). Milacic, M. et al. The Reactome Pathway Knowledgebase 2024. Nucleic Acids Research 52 , D672–D678 (2024). Agrawal, A. et al. WikiPathways 2024: next generation pathway database. Nucleic Acids Research 52 , D679–D689 (2024). Lambers Heerspink, H. J. & Gansevoort, R. T. Albuminuria Is an Appropriate Therapeutic Target in Patients with CKD: The Pro View. Clinical Journal of the American Society of Nephrology 10 , 1079 (2015). Zhao, J., Chang, L., Tu, J., Sun, B. & Wei, X. Evaluation of Annexins Family as Potential Biomarker for Predicting Progression and Prognosis in Clear Renal Cell Carcinoma. Evidence-Based Complementary and Alternative Medicine 2022 , 8748434 (2022). Shaikh, S. et al. Novel mutation in Cul7 gene in a family diagnosed with 3M syndrome. J Genet 98 , 21 (2019). Deeb, A., Afandi, O., Attia, S. & Fatih, A. E. 3-M syndrome: a novel CUL7 mutation associated with respiratory distress and a good response to GH therapy. Endocrinol Diabetes Metab Case Rep 2015 , 150012 (2015). Slieker, R. C. et al. HbA1c is associated with altered expression in blood of cell cycle- and immune response-related genes. Diabetologia 61 , 138–146 (2018). Fan, R. et al. Loss of the co-repressor GPS2 sensitizes macrophage activation upon metabolic stress induced by obesity and type 2 diabetes. Nat Med 22 , 780–791 (2016). Huo, Y., Lai, Y., Feng, Q., Wang, Q. & Li, J. Serum ITIH4 in coronary heart disease: a potential anti-inflammatory biomarker related to stenosis degree and risk of major adverse cardiovascular events. Biomark Med 16 , 1279–1288 (2022). Ma, Y. et al. ITIH4, as an inflammation biomarker, mainly increases in bacterial bloodstream infection. Cytokine 138 , 155377 (2021). Tian, L., Zhao, S. & Zhang, R. ITIH4 is a predictor for coronary thrombus in coronary arteriography patients. Future Cardiol 20 , 547–554 (2024). Ling, X.-B. et al. Mammalian Metallothionein-2A and Oxidative Stress. Int J Mol Sci 17 , 1483 (2016). Ściskalska, M., Ołdakowska, M. & Milnerowicz, H. Importance of Genetic Polymorphisms in MT1 and MT2 Genes in Metals Homeostasis and Their Relationship with the Risk of Acute Pancreatitis Occurrence in Smokers-Preliminary Findings. Int J Mol Sci 22 , 5725 (2021). Yang, C.-C. et al. The association of the blood lead level and serum lipid concentrations may be modified by the genetic combination of the metallothionein 2A polymorphisms rs10636 GC and rs28366003 AA. J Clin Lipidol 11 , 234–241 (2017). Karger, S. et al. Distinct pattern of oxidative DNA damage and DNA repair in follicular thyroid tumours. J Mol Endocrinol 48 , 193–202 (2012). Kim, H.-J. & Yun, J. Checkpoint-dependent phosphorylation of Med1/TRAP220 in response to DNA damage. Acta Biochim Biophys Sin (Shanghai) 49 , 496–502 (2017). Sperka, T., Wang, J. & Rudolph, K. L. DNA damage checkpoints in stem cells, ageing and cancer. Nat Rev Mol Cell Biol 13 , 579–590 (2012). Braun, D., Lelios, I., Krause, G. & Schweizer, U. Histidines in potential substrate recognition sites affect thyroid hormone transport by monocarboxylate transporter 8 (MCT8). Endocrinology 154 , 2553–2561 (2013). Filetti, S. & Rapoport, B. Inhibitors of specific aminoacyl-tRNA synthetases prevent thyrotropin-induced desensitization in cultured human thyroid cells. J Biol Chem 257 , 1342–1346 (1982). Kontaraki, J. E., Marketou, M. E., Zacharis, E. A., Parthenakis, F. I. & Vardas, P. E. MicroRNA-9 and microRNA-126 expression levels in patients with essential hypertension: potential markers of target-organ damage. J Am Soc Hypertens 8 , 368–375 (2014). Huang, Y. et al. Circulating miR-92a expression level in patients with essential hypertension: a potential marker of atherosclerosis. J Hum Hypertens 31 , 200–205 (2017). Kaleem, A. et al. Phosphorylated and O-GlcNAc Modified IRS-1 (Ser1101) and -2 (Ser1149) Contribute to Human Diabetes Type II. Protein Pept Lett 28 , 333–339 (2021). Ohtsubo, K. et al. N-Glycosylation modulates the membrane sub-domain distribution and activity of glucose transporter 2 in pancreatic beta cells. Biochem Biophys Res Commun 434 , 346–351 (2013). Gambineri, A. & Pelusi, C. Sex hormones, obesity and type 2 diabetes: is there a link? Endocr Connect 8 , R1–R9 (2019). Navarro, G. et al. Androgen excess in pancreatic β cells and neurons predisposes female mice to type 2 diabetes. JCI Insight 3 , e98607, 98607 (2018). Lubura, M. et al. Diabetes prevalence in NZO females depends on estrogen action on liver fat content. Am J Physiol Endocrinol Metab 309 , E968-980 (2015). Brandt, C. et al. Plasma and muscle myostatin in relation to type 2 diabetes. PLoS One 7 , e37236 (2012). Gahete, M. D. et al. Changes in Splicing Machinery Components Influence, Precede, and Early Predict the Development of Type 2 Diabetes: From the CORDIOPREV Study. EBioMedicine 37 , 356–365 (2018). Kumar, P., Gupta, R. & Gupta, A. Vitamin D deficiency in patients with diabetes and its correlation with water fluoride levels. J Water Health 21 , 125–137 (2023). Lobo, J. G. V. M. et al. Low-Level Fluoride Exposure Increases Insulin Sensitivity in Experimental Diabetes. J Dent Res 94 , 990–997 (2015). Bagley, J., Yuan, J., Chandrakar, A. & Iacomini, J. Hyperlipidemia Alters Regulatory T Cell Function and Promotes Resistance to Tolerance Induction Through Costimulatory Molecule Blockade. Am J Transplant 15 , 2324–2335 (2015). Chen, S. et al. Hyperlipidemia modifies innate immune responses to lipopolysaccharide via the TLR-NF-κB signaling pathway. Inflammation 36 , 968–976 (2013). Mehta, S. R., Yusuf, S., & Clopidogrel in Unstable angina to prevent Recurrent Events (CURE) Study Investigators. The Clopidogrel in Unstable angina to prevent Recurrent Events (CURE) trial programme; rationale, design and baseline characteristics including a meta-analysis of the effects of thienopyridines in vascular disease. Eur Heart J 21 , 2033–2041 (2000). Savi, P. & Herbert, J.-M. Clopidogrel and Ticlopidine: P2Y 12 Adenosine Diphosphate-Receptor Antagonists for the Prevention of Atherothrombosis. Semin Thromb Hemost 31 , 174–183 (2005). Cheng, Z. et al. Gene expression profile-based drug screen identifies SAHA as a novel treatment for NAFLD. Mol. Omics 15 , 50–58 (2019). Steffens, J. P. et al. Role of testosterone and androgen receptor in periodontal disease progression in female rats. Journal of Periodontology 91 , 545–553 (2020). Shapiro, L. F. & Freeman, K. The relationship between estrogen, estrogen receptors and periodontal disease in adult women. J Mich Dent Assoc 96 , 40–44 (2014). Santana, L. C. L. et al. Testosterone Increases Fibroblast Proliferation in vitro Through Androgen and Estrogen Receptor Activation. J Int Acad Periodontol 22 , 146–155 (2020). Fizazi, K. et al. Rucaparib or Physician’s Choice in Metastatic Prostate Cancer. N Engl J Med 388 , 719–732 (2023). Saha, L. Role of peroxisome proliferator-activated receptors alpha and gamma in gastric ulcer: An overview of experimental evidences. WJGPT 6 , 120 (2015). Verdrengh, M., Isaksson, O. & Tarkowski, A. Topoisomerase II inhibitors, irrespective of their chemical composition, ameliorate experimental arthritis. Rheumatology 44 , 183–186 (2005). Napoleone, E. et al. Inhibition of the renin-angiotensin system downregulates tissue factor and vascular endothelial growth factor in human breast carcinoma cells. Thrombosis Research 129 , 736–742 (2012). Sato, T., Li, Y., Saito, T. & Nakaya, H. Minoxidil opens mitochondrial KATP channels and confers cardioprotection. British Journal of Pharmacology 141 , 360–366 (2004). Nichols, C. G., Singh, G. K. & Grange, D. K. KATP Channels and Cardiovascular Disease. Circulation Research 112 , 1059–1072 (2013). Van Zanten, A. R. H. et al. Continuous vs. intermittent cefotaxime administration in patients with chronic obstructive pulmonary disease and respiratory tract infections: pharmacokinetics/pharmacodynamics, bacterial susceptibility and clinical efficacy. Brit J Clinical Pharma 63 , 100–109 (2007). Li, H. et al. Immune regulation by low doses of the DNA methyltransferase inhibitor 5-azacitidine in common human epithelial cancers. Oncotarget 5 , 587–598 (2014). Hsi, L. C., Xi, X., Wu, Y. & Lippman, S. M. The methyltransferase inhibitor 5-aza-2-deoxycytidine induces apoptosis via induction of 15-lipoxygenase-1 in colorectal cancer cells. Molecular Cancer Therapeutics 4 , 1740–1746 (2005). Li, Y., Fu, J., Yuan, X. & Hu, C. Simvastatin inhibits the proliferation of A549 lung cancer cells through oxidative stress and up-regulation of SOD2. Pharmazie 69 , 610–614 (2014). Liu, H., Wang, Z., Li, Y., Li, W. & Chen, Y. Simvastatin prevents proliferation and bone metastases of lung adenocarcinoma in vitro and in vivo. neo 60 , 240–246 (2013). Yao, J., Zhang, Y., Ramishetti, S., Wang, Y. & Huang, L. Turning an antiviral into an anticancer drug: Nanoparticle delivery of acyclovir monophosphate. Journal of Controlled Release 170 , 414–420 (2013). Pezaro, C. PARP inhibitor combinations in prostate cancer. Ther Adv Med Oncol 12 , 1758835919897537 (2020). Ververis, K. & Karagiannis, T. C. Potential non-oncological applications of histone deacetylase inhibitors. Am J Transl Res 3 , 454–467 (2011). Wu, C. W. et al. Regulation of arginase production by glucocorticoid in three human gastric cancer cell lines. Life Sci 51 , 1355–1361 (1992). Busada, J. T. et al. Endogenous glucocorticoids prevent gastric metaplasia by suppressing spontaneous inflammation. J Clin Invest 129 , 1345–1358 (2019). Tanaka, A. et al. Role of copper ion in the pathogenesis of type 2 diabetes. Endocr J 56 , 699–706 (2009). Nagai, R., Murray, D. B., Metz, T. O. & Baynes, J. W. Chelation: a fundamental mechanism of action of AGE inhibitors, AGE breakers, and other inhibitors of diabetes complications. Diabetes 61 , 549–559 (2012). Chenoufi, N. et al. Antiproliferative effect of deferiprone on the Hep G2 cell line. Biochem Pharmacol 56 , 431–437 (1998). Kontoghiorghes, G. J., Efstathiou, A., Ioannou-Loucaides, S. & Kolnagou, A. Chelators controlling metal metabolism and toxicity pathways: applications in cancer prevention, diagnosis and treatment. Hemoglobin 32 , 217–227 (2008). Hawash, M. Recent Advances of Tubulin Inhibitors Targeting the Colchicine Binding Site for Cancer Therapy. Biomolecules 12 , 1843 (2022). Peerzada, M. N., Dar, M. S. & Verma, S. Development of tubulin polymerization inhibitors as anticancer agents. Expert Opin Ther Pat 33 , 797–820 (2023). Dimas, K. et al. Steroidal cardiac Na+/K+ ATPase inhibitors exhibit strong anti-cancer potential in vitro and in prostate and lung cancer xenografts in vivo. Anticancer Agents Med Chem 14 , 762–770 (2014). Joshi, S. K., Mikusa, J. P., Weaver, B. & Honore, P. Morphine and ABT-594 (a nicotinic acetylcholine agonist) exert centrally mediated antinociception in the rat cyclophosphamide cystitis model of visceral pain. J Pain 9 , 146–156 (2008). Palea, S., Artibani, W., Ostardo, E., Trist, D. G. & Pietra, C. Evidence for purinergic neurotransmission in human urinary bladder affected by interstitial cystitis. J Urol 150 , 2007–2012 (1993). Gosis, B. S. et al. Inhibition of nonalcoholic fatty liver disease in mice by selective inhibition of mTORC1. Science 376 , eabf8271 (2022). Feng, J. et al. mTOR: A Potential New Target in Nonalcoholic Fatty Liver Disease. Int J Mol Sci 23 , 9196 (2022). Deng, T. et al. Comparison of Genotype Imputation for SNP Array and Low-Coverage Whole-Genome Sequencing Data. Front Genet 12 , 704118 (2021). An, J. et al. Blood transcriptome differentiates clinical clusters for asthma. World Allergy Organ J 17 , 100871 (2024). Lee, J. & Gim, J. Unlocking the Potential of Low Quality Total RNA-seq Data: A Stepwise Mapping Approach for Improved Quantitative Analyses. 2024.09.30.615750 Preprint at https://doi.org/10.1101/2024.09.30.615750 (2024). Moon, S. et al. The Korea Biobank Array: Design and Identification of Coding Variants Associated with Blood Biochemical Traits. Sci Rep 9 , 1382 (2019). Seo, S. et al. SNP genotype calling and quality control for multi-batch-based studies. Genes Genomics 41 , 927–939 (2019). Song, Y. E. et al. ONETOOL for the analysis of family-based big data. Bioinformatics 34 , 2851–2853 (2018). Howie, B. N., Donnelly, P. & Marchini, J. A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. PLoS Genet 5 , e1000529 (2009). Yoo, S.-K. et al. NARD: whole-genome reference panel of 1779 Northeast Asians improves imputation accuracy of rare and low-frequency variants. Genome Med 11 , 64 (2019). Loh, P.-R. et al. Reference-based phasing using the Haplotype Reference Consortium panel. Nat Genet 48 , 1443–1448 (2016). Fuchsberger, C., Abecasis, G. R. & Hinds, D. A. minimac2: faster genotype imputation. Bioinformatics 31 , 782–784 (2015). Bolger, A. M., Lohse, M. & Usadel, B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30 , 2114–2120 (2014). Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29 , 15–21 (2013). DeLuca, D. S. et al. RNA-SeQC: RNA-seq metrics for quality control and process optimization. Bioinformatics 28 , 1530–1532 (2012). Harrow, J. et al. GENCODE: the reference human genome annotation for The ENCODE Project. Genome Res 22 , 1760–1774 (2012). Robinson, M. D. & Oshlack, A. A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol 11 , R25 (2010). Stegle, O., Parts, L., Piipari, M., Winn, J. & Durbin, R. Using probabilistic estimation of expression residuals (PEER) to obtain increased power and interpretability of gene expression analyses. Nat Protoc 7 , 500–507 (2012). GTEx Consortium et al. Genetic effects on gene expression across human tissues. Nature 550 , 204–213 (2017). Friedman, J., Hastie, T. & Tibshirani, R. Regularization Paths for Generalized Linear Models via Coordinate Descent. J Stat Softw 33 , 1–22 (2010). Shabalin, A. A. Matrix eQTL: ultra fast eQTL analysis via large matrix operations. Bioinformatics 28 , 1353–1358 (2012). Yang, J., Lee, S. H., Goddard, M. E. & Visscher, P. M. GCTA: a tool for genome-wide complex trait analysis. Am J Hum Genet 88 , 76–82 (2011). Kuznetsova, I., Lugmayr, A., Siira, S. J., Rackham, O. & Filipovska, A. CirGO: an alternative circular way of visualising gene ontology terms. BMC Bioinformatics 20 , 84 (2019). Knox, C. et al. DrugBank 6.0: the DrugBank Knowledgebase for 2024. Nucleic Acids Res 52 , D1265–D1275 (2024). Additional Declarations There is NO Competing Interest. Supplementary Files 20250410SupplyFigures.pdf All Supplementary Figure File 20250410SupplyTables.xlsx All Supplementary Table File Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6417034","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":443880373,"identity":"c6ec3d49-be44-455b-a06a-68b699236af6","order_by":0,"name":"Sungho Won","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAo0lEQVRIiWNgGAWjYJCCAx+gDMYGInUwHpxBqhbmwzwkaTE43nvgsM2fO4kN7IcfMM7cQ4yWM+cSDue2PUts4EkzYNzwjAgtZjdyDA7nNhxObGDIYWB8cIBYLRZ/gFr435CihYENqEUCaMsGYrTYnzljcLC37bBxm8Qzg4MziNEi2d5j/OHHn8Oy/fzJDx/2EKMFDtiAmCQNo2AUjIJRMArwAAAUVz3KJtw6TwAAAABJRU5ErkJggg==","orcid":"","institution":"Seoul National University","correspondingAuthor":true,"prefix":"","firstName":"Sungho","middleName":"","lastName":"Won","suffix":""},{"id":443880374,"identity":"f2c715f4-75dd-41ea-bb3a-6d7eb2365cad","order_by":1,"name":"Min Heo","email":"","orcid":"https://orcid.org/0000-0002-4217-9358","institution":"Seoul National University","correspondingAuthor":false,"prefix":"","firstName":"Min","middleName":"","lastName":"Heo","suffix":""},{"id":443880375,"identity":"7a4c1f1e-3f33-42b4-9484-c1a6840be076","order_by":2,"name":"Yunmi Ji","email":"","orcid":"https://orcid.org/0009-0002-7914-1625","institution":"Seoul National University","correspondingAuthor":false,"prefix":"","firstName":"Yunmi","middleName":"","lastName":"Ji","suffix":""},{"id":443880376,"identity":"7b58f91b-cbff-4fbc-9d20-fd7c66bd699a","order_by":3,"name":"Ah Ra Do","email":"","orcid":"","institution":"Seoul National University","correspondingAuthor":false,"prefix":"","firstName":"Ah","middleName":"Ra","lastName":"Do","suffix":""},{"id":443880377,"identity":"c3d8e245-e299-4d6d-aa77-f09c1a2f1311","order_by":4,"name":"Heejin Jin","email":"","orcid":"","institution":"Seoul National University","correspondingAuthor":false,"prefix":"","firstName":"Heejin","middleName":"","lastName":"Jin","suffix":""},{"id":443880378,"identity":"f188b33f-32ef-4149-9fda-c03e39f81a2e","order_by":5,"name":"Kyungtaek Park","email":"","orcid":"","institution":"Interdisciplinary Program of Bioinformatics,Seoul National University","correspondingAuthor":false,"prefix":"","firstName":"Kyungtaek","middleName":"","lastName":"Park","suffix":""},{"id":443880379,"identity":"fd9e39f5-2908-459e-ae8f-be3668444b55","order_by":6,"name":"Nam-Eun Kim","email":"","orcid":"https://orcid.org/0000-0003-4155-2584","institution":"Seoul National University","correspondingAuthor":false,"prefix":"","firstName":"Nam-Eun","middleName":"","lastName":"Kim","suffix":""},{"id":443880380,"identity":"25ea5ae4-e964-42c6-be7e-eac176aa5ba2","order_by":7,"name":"Juhee Ahn","email":"","orcid":"","institution":"Institute of Health and Environment, Seoul National Univeristy","correspondingAuthor":false,"prefix":"","firstName":"Juhee","middleName":"","lastName":"Ahn","suffix":""},{"id":443880381,"identity":"8ea5b799-6393-42f4-9767-50f55afd1623","order_by":8,"name":"Tae-Bum Kim","email":"","orcid":"","institution":"Asan Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Tae-Bum","middleName":"","lastName":"Kim","suffix":""},{"id":443880382,"identity":"7bc53f1a-bfdd-49be-8b31-bcebc2a3497c","order_by":9,"name":"Kun Ho Lee","email":"","orcid":"https://orcid.org/0000-0003-4144-8162","institution":"Chosun University","correspondingAuthor":false,"prefix":"","firstName":"Kun","middleName":"Ho","lastName":"Lee","suffix":""},{"id":443880383,"identity":"640f8276-1ca8-42b7-9529-7ab259a71016","order_by":10,"name":"Jungsoo Gim","email":"","orcid":"https://orcid.org/0000-0002-6454-9751","institution":"Department of Biomedical Science, Chosun University","correspondingAuthor":false,"prefix":"","firstName":"Jungsoo","middleName":"","lastName":"Gim","suffix":""},{"id":443880384,"identity":"b0e6a8e9-78a1-4eb9-979e-b19cd32c66e2","order_by":11,"name":"Stefan Konigorski","email":"","orcid":"https://orcid.org/0000-0002-9966-6819","institution":"Hasso Plattner Institute for Digital Engineering","correspondingAuthor":false,"prefix":"","firstName":"Stefan","middleName":"","lastName":"Konigorski","suffix":""}],"badges":[],"createdAt":"2025-04-10 06:35:34","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6417034/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6417034/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":80806097,"identity":"c630733f-12bb-4433-bd98-d0c7c4c98102","added_by":"auto","created_at":"2025-04-17 09:28:27","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":561719,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOverview of the study. \u003c/strong\u003eGReX; Genetically Regulated gene eXpression\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6417034/v1/1c8f8c8bbd314173dfccaff5.png"},{"id":80806095,"identity":"45297f9b-6420-4ce2-9a23-3f93e3151ebc","added_by":"auto","created_at":"2025-04-17 09:28:27","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":613440,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOverall comparison between the GTEx-based GReX model and the Korean-based GReX model. \u003c/strong\u003e(A) Venn-diagram of genes in each model, (B) histogram and marginal plot for R-squared (𝑅\u003csup\u003e2\u003c/sup\u003e) between predicted gene expression and real gene expression, (C) histogram of the SNP heritability (ℎ\u003csup\u003e2\u003c/sup\u003e) of the expression of genes in the GReX models\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6417034/v1/c3e71073eaf023a33e50d577.png"},{"id":80806096,"identity":"07bbfc23-dbcb-418a-8bd0-93a7a42dbee9","added_by":"auto","created_at":"2025-04-17 09:28:27","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1579727,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAll significant TWAS results. \u003c/strong\u003eTWAS results of 81 traits based on both Korean and GTEx-based models. Only gene-trait associations satisfying the Bonferroni adjusted \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05 were shown. The shape was separated by model case and the color was separated by phenotypes.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6417034/v1/a984d0828a963bcf8cac1706.png"},{"id":80806102,"identity":"6f4b181f-2d52-4a1b-8ade-0b242cbd49cf","added_by":"auto","created_at":"2025-04-17 09:28:27","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2437516,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGO enrichment analysis. \u003c/strong\u003eThe Biological Processes (BP) in the GO terms were identified using MSigDB and further refined using REVIGO, which uses internal algorithms to cluster and summarise the terms. The results were visualised using CirGO, which displays the proportion of genes associated with each GO category in a pie chart. The proportions of the clusters are indicated in the legend, and the subset of GO terms grouped under each representative category are shown as bar graphs, highlighting their significance with –log\u003csub\u003e10\u003c/sub\u003e(FDR) values for each ontology. (A) Thyroid disease-Korean-based model, (B) Hypertension-GTEx-based model, (C) Top: Type 2 Diabetes-Korean-based model, Bottom: Type 2 Diabetes-GTEx-based model, (D) Top: Hyperlipidemia-Korean-based model, Bottom: Hyperlipidemia-GTEx-based model\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6417034/v1/3688cca8159439c947b31017.png"},{"id":84829491,"identity":"d27f5ba5-4d32-4717-95a7-ef011e0d9b4f","added_by":"auto","created_at":"2025-06-17 18:28:31","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":15005298,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6417034/v1/a5b314a7-a8e5-4be3-a8c6-7f5fc74ab9c3.pdf"},{"id":80806093,"identity":"bd3606fd-650c-4d4f-9554-7044a70e372f","added_by":"auto","created_at":"2025-04-17 09:28:27","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":583396,"visible":true,"origin":"","legend":"All Supplementary Figure File","description":"","filename":"20250410SupplyFigures.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6417034/v1/378624fdb89fff45c3eb6973.pdf"},{"id":80806094,"identity":"bda26927-f25e-4696-a50a-819a09b9610c","added_by":"auto","created_at":"2025-04-17 09:28:27","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":1986453,"visible":true,"origin":"","legend":"All Supplementary Table File","description":"","filename":"20250410SupplyTables.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6417034/v1/b00d093c5daf03cbe0b8a2e2.xlsx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Transcriptome-Wide Association Studies of 81 traits in 79,294 Korean individuals","fulltext":[{"header":"Introduction","content":"\u003cp\u003eGenome-wide association studies (GWAS) have successfully identified numerous genetic loci associated with complex traits. However, interpreting these loci remains challenging due to linkage disequilibrium (LD), which often obscures the causal variants driving these associations. Moreover, GWAS alone typically do not allow to pinpoint the target genes mediating the effects of causal variants on traits, and are constrained by the burden of multiple testing\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. To address these interpretability and statistical power issues, transcriptome-wide association studies (TWAS) have emerged as a powerful approach in large-scale association studies\u003csup\u003e\u003cspan additionalcitationids=\"CR3\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eTWAS integrate genetic data with gene expression levels, using genetically regulated gene expression (GReX) to identify risk genes associated with complex traits and diseases. The standard two-stage TWAS approach involves first constructing gene expression imputation models using a transcriptomic reference panel, where gene expression is treated as the response and genetic variants are used as predictors. Associated variants are called expression quantitative trait loci(eQTLs) their estimated effect sizes are then used as weights in gene-based association tests with GWAS data. By incorporating eQTLs from relevant tissues, TWAS improves interpretability and identifies functionally relevant loci. This gene-centric method consolidates the effects of various regulatory variants into a singular test unit, enhancing the study's power and yielding more interpretable genomic loci associated with traits\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eDespite its advantages, TWAS reference panels, such as the widely used panel of the Genotype-Tissue Expression (GTEx) project, are predominantly derived from individuals of European ancestry. This lack of diversity limits the applicability of TWAS to other populations\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e, as population-specific differences in allele frequencies, LD patterns, and haplotypes can reduce the accuracy of genetic signal detection, leading to false negatives in non-European populations\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. This bias has hindered the comprehensive understanding of complex traits across diverse populations and ancestries. Recent studies underscore the importance of incorporating ancestry-specific models to improve accuracy in genetic association studies across global populations\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn this study, we developed a reference panel specifically for the Korean populations and compared its performance with the GTEx panel, which is based predominantly on European individuals. Using these panels, we conducted a large-scale TWAS on 81 traits in 79,294 Korean individuals. Our analysis identified 348 significant gene-trait associations (Bonferroni-corrected \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), including 232 novel associations using the Korean panel, compared to 180 associations using the GTEx panel. Furthermore, using these significant gene-trait associations, pathway enrichment analysis revealed distinct biological processes that were enriched in the Korean-based model, such as thyroid disease, type 2 diabetes, and hyperlipidaemia that were distinct from those detected using the GTEx panel. Additionally, computational drug repurposing prioritized promising therapeutic candidates, such as clopidogrel for angina and vorinostat for fatty liver disease, highlighting the translational potential of these findings. These results underscore the importance of population-specific approaches in uncovering novel genetic associations and provide critical insights into the genetic architecture of complex traits in underrepresented populations. Our study emphasizes the necessity of diverse reference panels to advance precision medicine and improve global health equity.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eOverview of the Study\u003c/h2\u003e \u003cp\u003eIn this study, we performed GReX modeling by integrating transcriptomic data extracted from whole blood tissue in two Korean cohorts (Asan and Chosun cohorts). We then compared the results from Korean-based GReX model with the GTEx-based GReX model, which was primarily derived from European individuals. Next, we applied both models to genotype data from the Korean Genome and Epidemiology Study (KoGES)\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e and and the Gene-environmental interaction and phenotype (GENIE)\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e cohorts (total N\u0026thinsp;=\u0026thinsp;79,294) and conducted TWAS for 81 phenotypes. To evaluate the novelty and biological relevance of the TWAS-identified gene-trait associations, we utilized the Harmonizome\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e database and conducted pathway enrichment analyses\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Finally, we employed the CMap linked user environment (CLUE)\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e for computational drug repurposing to identify bioactive small molecules capable of reversing the expression profiles of clinically relevant trait-associated genes\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eComparison of GReX Models\u003c/h3\u003e\n\u003cp\u003eWe focused on genes meeting two criteria: an average Pearson correlation coefficient between predicted and observed gene expression\u0026thinsp;\u0026gt;\u0026thinsp;0.1 and \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 in the association test between predicted and observed gene expression based on Stouffer\u0026rsquo;s method (see Methods). The GTEx-based model included 7,252 genes, whereas the Korean-based model comprised 4,699 genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). When evaluating performance on the respective training datasets, the GTEx-based model outperformed the Korean-based model (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}=\\)\u003c/span\u003e\u003c/span\u003e0.11 vs \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}=\\)\u003c/span\u003e\u003c/span\u003e0.07). However, when evaluated on the Korean data, the GTEx-based model underperformed compared to the Korean-based model (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}=\\)\u003c/span\u003e\u003c/span\u003e0.03 vs. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}=\\)\u003c/span\u003e\u003c/span\u003e0.07) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). This trend held for intersecting genes, with the Korean-based model showing better performance on Korean data (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}=\\)\u003c/span\u003e\u003c/span\u003e0.08 vs \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}=\\)\u003c/span\u003e\u003c/span\u003e0.06) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo further evaluate the models, we calculated the cis-heritability of the genes in both model on the Korean data. The mean of cis-heritability for genes in the Korean-based model was higher compared to the GTEx-based model (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{h}^{2}=\\)\u003c/span\u003e\u003c/span\u003e0.09 vs \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{h}^{2}=\\)\u003c/span\u003e\u003c/span\u003e0.05). Notably, the genes common to both models exhibited an even higher mean cis-heritability (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{h}^{2}=\\)\u003c/span\u003e\u003c/span\u003e0.11) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC, Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn addition, allele frequencies of SNPs differed by population: SNPs in the Korean-based model had lower frequencies in East Asians, whereas those in the GTEx-based model had lower frequencies in Europeans\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). An eQTL analysis revealed that 94% of genes in the Korean-based model (4,434/4,699) were significantly associated with SNPs (FDR-adjusted \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), compared to 49% (3,553/7,252) in the GTEx-based model. Similarly, the Korean-based model had a higher proportion of significant eQTL-gene pairs (54.8% vs 19.3%) (Table S2).\u003c/p\u003e\n\u003ch3\u003eAssociations of Predicted Gene Expression with 81 Traits\u003c/h3\u003e\n\u003cp\u003eOf the 81,153 individuals in KoGES\u0026thinsp;+\u0026thinsp;GENIE cohorts, 79,294 with no missing data for age, sex, BMI, and smoking status were considered for TWAS for 81 traits. We assessed both the Korean-based and GTEx-based models, finding that their correlations of predicted gene expression obtained from intersecting genes was relatively high (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}=\\)\u003c/span\u003e\u003c/span\u003e0.40), with most correlations being positive (Figure S2). TWAS was then conducted for genes with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}\\)\u003c/span\u003e\u003c/span\u003e (see Methods) greater than 0.05, yielding 1,104 genes from the GTEx-based model (out of 7,252 total) and 1,630 genes from the Korean-based model (out of 4,699 total). Using a Bonferroni-adjusted \u003cem\u003eP\u003c/em\u003e threshold, the Korean-based model identified 348 significant gene-trait associations (181 genes across 37 traits), whereas the GTEx-based model identified 260 (139 genes across 34 traits). Among these, 138 associations (72 genes across 31 traits) were significant in both models (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, Table S7). Notably, the Korean-based model identified more significant genes than the GTEx-based model in most traits (33 out of 39 traits; Figure S3). Detailed trait information is provided in Table S3.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe also examined the significance and directionality of TWAS \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:Z\\)\u003c/span\u003e\u003c/span\u003e-scores for gene-trait associations common to both models. Most significant gene-trait associations were positive and unique to the Korean-based model (not significant in GTEx-based model), followed by negative associations unique to the Korean-based model. Additionally, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:Z\\)\u003c/span\u003e\u003c/span\u003e-scores for overlapping gene-trait associations showed consistent directionality (Figure S4), with a correlation of 0.4229 across two models (Figure S5).\u003c/p\u003e\n\u003ch3\u003eIdentification of Novel Gene-Trait Associations\u003c/h3\u003e\n\u003cp\u003eUsing the Harmonizome database\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e, we found that 137 gene-trait associations unique to the Korean-based model were novel (44 strong, 55 weak, 38 no). In contrast, the GTEx-based model identified 85 novel associations (22 strong, 17 weak, 46 no). Among shared associations, 95 were novel (32 strong, 23 weak, and 40 no) (Table S5, Table S6).\u003c/p\u003e\n\u003ch3\u003ePathway Enrichment Analysis\u003c/h3\u003e\n\u003cp\u003eFor diseases associated with at least three genes, we performed a gene ontology (GO) analysis. GO analysis identified 369 enriched biological processes using the Korean-based model, and 294 using the GTEx-based model (FDR \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.1). Clustering analysis with REVIGO\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e highlighted distinct biological patterns.\u003c/p\u003e \u003cp\u003eBoth models identified pathways for type 2 diabetes and hyperlipidemia, but with distinct enriched terms (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, Table S8). For type 2 diabetes, the Korean-based model highlighted terms such as \u0026lsquo;O-glycan processing\u0026rsquo; (34.6%), \u0026lsquo;female gonad development\u0026rsquo; (18.1%), \u0026lsquo;negative regulation of muscle hypertrophy\u0026rsquo; (11.5%), and \u0026lsquo;natural killer cell-mediated cytotoxicity directed against tumor cell targets\u0026rsquo; (11.5%). In contrast, the GTEx-based model uncovered distinct terms, including \u0026lsquo;RNA splicing\u0026rsquo; (27.6%), \u0026lsquo;negative regulation of antibody-dependent cellular cytotoxicity\u0026rsquo; (19.1%), \u0026lsquo;response to fluoride\u0026rsquo; (18.1%), and \u0026lsquo;specification of segmental identity antennal segment\u0026rsquo; (14.1%). Notably, no overlapping clusters were identified between the two models for type 2 diabetes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC). Similarly, both models produced significant results for hyperlipidemia, but again with non-overlapping enriched terms. The Korean-based model identified terms such as \u0026lsquo;endonucleolytic cleavage to generate mutated 5\u0026rsquo;-end of SSU-rRNA\u0026rsquo; (26.0%), \u0026lsquo;female meiosis chromosome segregation\u0026rsquo; (17.7%), \u0026lsquo;nucleobase-containing compound transport\u0026rsquo; (13.4%), and \u0026lsquo;regulation of response to biotic stimulus\u0026rsquo; (12.8%). In contrast, the GTEx-based model identified different terms such as \u0026lsquo;regulation of biosynthetic process of antibacterial peptides active against Gram-positive bacteria\u0026rsquo; (37.2%), \u0026lsquo;synaptic vesicle targeting\u0026rsquo; (22.0%), \u0026lsquo;enzyme-directed rRNA pseudouridine synthesis\u0026rsquo; (19.6%), and \u0026lsquo;inflorescence meristem growth\u0026rsquo; (12.8%) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn addition, significant terms for thyroid disease were identified only by the Korean-based model, whereas hypertension terms were exclusive to the GTEx-based model. For thyroid disease, the top enriched GO terms in the Korean-based model included \u0026lsquo;auxin biosynthetic process\u0026rsquo; (40.1%), \u0026lsquo;negative regulation of DNA damage checkpoint\u0026rsquo; (18.7%), \u0026lsquo;inflorescence meristem growth\u0026rsquo; (10.7%), and \u0026lsquo;histidine transport\u0026rsquo; (10.1%) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). In contrast, hypertension-related terms in the GTEx-based model involved terms such as somatic diversification of immunoglobulins\u0026rsquo; (60.1%), \u0026lsquo;auxin export across the plasma membrane\u0026rsquo; (14.0%), \u0026lsquo;negative regulation of atrichoblast fate specification\u0026rsquo; (7.1%), and \u0026lsquo;positive regulation of post-transcriptional gene silencing\u0026rsquo; (10.1%) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). Detailed pathway enrichment results, including GO\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, KEGG\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, Reactome\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e, WikiPathways\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e analyses are provided in Table S8.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eComputational Drug Repurposing Analysis\u003c/h2\u003e \u003cp\u003eWe prioritized drug candidates with therapeutic potential using Connectivity Map (CMap) database derived from the L1000 assay\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Using the Korean-based model, we identified significant drug-disease associations. Clopidogrel was associated with angina, vorinostat with fatty liver disease, fulvestrant with periodontal disease, rucaparib with prostate cancer, and pioglitazone with gastritis. Similarly, the GTEx-based model revealed additional significant drug-disease associations. Etoposide was linked to arthritis, enalapril to breast cancer, minoxidil to heart failure, cefotaxime to chronic pulmonary disease, azacitidine to colorectal cancer, simvastatin to lung cancer, and acyclovir to prostate cancer. Only the top-ranked (top 1) drug for each disease was subjected to a literature review (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) and the comprehensive list of identified drug-disease pairs is provided in Table S9.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e \u003cb\u003eTop 1 drug repurposing results.\u003c/b\u003e Drug enrichment analysis of TWAS results indicates biologically relevant drugs. We imported drug target genes from CLUE and tested whether these gene sets were enriched by TWAS hits. We only displayed Top 1 drug repurposing results per phenotype that satisfied a connectivity score (tau) \u0026lt;= -90 according to each model. All significant results can be found in Table S9.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePhenotype\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDrug name\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMechanism of Action\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eKorean_Tau\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGTEx_Tau\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"24\" rowspan=\"25\"\u003e \u003cp\u003eKorean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eangina\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eclopidogrel\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003epurinergic receptor antagonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-91.02\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-88.45\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003easthma\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003epanobinostat\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eHDAC inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-98.10\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eprostate hyperplasia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eprednicarbate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ephospholipase activator\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-96.69\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-9.57\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003echronic pulmonary disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ebifonazole\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003esterol demethylase inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-95.38\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-1.93\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecolorectal cancer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003edigoxin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eATPase inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-94.16\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-70.19\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etype 2 diabetes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eflunisolide\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ecytochrome P450 inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-95.33\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efatty liver\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003evorinostat\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eHDAC inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-98.29\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-2.68\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efracture\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003etivozanib\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eVEGFR inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-94.623\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e11.27\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003echronic gastritis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eetodolac\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ecyclooxygenase inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-94.41\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-4.86\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003egallstones\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003epitavastatin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eHMGCR inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-94.51\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-10.39\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003egastric cancer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003etriamcinolone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eglucocorticoid receptor agonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-90.62\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-80.08\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eliver cancer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003edeferiprone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003echelating agent\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-94.70\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-68.52\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003echronic hepatitis C\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003epanobinostat\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eHDAC inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-95.17\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e9.74\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ehypertension\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eselumetinib\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMEK inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-94.54\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-0.14\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ehyperlipidemia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eclotrimazole\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ecytochrome P450 inhibitor|imidazoline receptor ligand\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-90.67\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-74.17\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eacute liver disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003epropofol\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ebenzodiazepine receptor agonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-94.72\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-11.05\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003emyocardial infarction\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eoxiconazole\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ebacterial cell wall synthesis inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-93.57\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e8.25\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eperiodontal disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003efulvestrant\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eestrogen receptor antagonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-95.77\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-96.23\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eprostate cancer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003erucaparib\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePARP inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-92.87\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-1.76\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003egastritis, gastric/duodenal ulcer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003epioglitazone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003einsulin sensitizer|PPAR receptor agonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-91.22\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-2.08\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etuberculosis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003enoscapine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ebradykinin receptor antagonist|tubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-90.30\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-31.46\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ethyroid cancer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003edigoxin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eATPase inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-97.74\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etransient ischemic attack (TIA)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003efulvestrant\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eestrogen receptor antagonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-95.02\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-18.94\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecystitis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ehomatropine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eacetylcholine receptor antagonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-90.02\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e18.86\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003evalvular disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ealbendazole\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-93.80\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-0.18\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"34\" rowspan=\"35\"\u003e \u003cp\u003eGTEx\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eangina\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003edocetaxel\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e95.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-90.37\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003earrhythmia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003evinorelbine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-17.90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-95.88\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003earthritis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eetoposide\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etopoisomerase inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-32.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-94.08\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003easthma\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003evinorelbine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-73.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-94.49\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ebreast cancer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eenalapril\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eangiotensin converting enzyme inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-10.27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-90.38\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecoronary artery disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003epropranolol\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eadrenergic receptor antagonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-6.20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-92.49\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecataract\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eliothyronine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ethyroid hormone stimulant\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-73.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-96.70\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eheart failure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eminoxidil\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eKATP activator|Kir6 channel (KATP) activator|vasodilator\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-13.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-99.15\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eliver cirrhosis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003efluvoxamine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eselective serotonin reuptake inhibitor (SSRI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-98.22\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003echronic pulmonary disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ecefotaxime\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ebacterial cell wall synthesis inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-4.48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-91.80\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecolorectal cancer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eazacitidine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDNA methyltransferase inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-93.69\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecerebrovascular disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eolanzapine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003edopamine receptor antagonist|serotonin receptor antagonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-92.42\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etype 2 diabetes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003edeferiprone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003echelating agent\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-45.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-99.12\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efatty liver\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003esirolimus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003emTOR inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-2.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-91.95\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efracture\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003emetformin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003einsulin sensitizer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-96.25\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003echronic gastritis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003evinorelbine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-74.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-93.15\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003egallstones\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003esulpiride\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003edopamine receptor antagonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-44.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-95.08\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003egout\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003evinorelbine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-69.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-94.90\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003echronic hepatitis B\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003evidarabine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eantiviral\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-4.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-97.11\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ehypertension\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003evinorelbine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-85.45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-95.87\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ekidney disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003elypressin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003evasopressin receptor agonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-94.44\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003elung cancer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003esimvastatin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eHMGCR inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-95.10\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ehyperlipidemia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003edocetaxel\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-39.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-90.12\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eacute liver disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ediclofenamide\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ecarbonic anhydrase inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-0.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-93.92\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003emyocardial infarction\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003erifabutin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eprotein synthesis inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-8.09\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-91.87\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eperiodontal disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eeugenol\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eandrogen receptor antagonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-13.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-97.65\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eprostate cancer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eacyclovir\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDNA polymerase inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-92.25\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eperipheral vascular disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003evinorelbine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-6.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-90.26\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003esmoking status\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003edocetaxel\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-87.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-90.30\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etuberculosis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003edeferiprone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003echelating agent\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-44.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-92.63\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ethyroid disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ealbendazole\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e15.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-90.60\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ethyroid cancer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003evincristine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003etubulin polymerization inhibitor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-13.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-91.21\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etransient ischemic attack\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eoxybutynin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eacetylcholine receptor antagonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e15.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-98.17\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecystitis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003evecuronium\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eacetylcholine receptor antagonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-4.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-99.89\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003evalvular disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ebenperidol\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003edopamine receptor antagonist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-14.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-92.25\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003eKorean_Tau: Connectivity score of durg repurposing using genes with significant association with each phenotype when TWAS was performed based on Korean-based model\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003eGTEx_Tau: Connectivity score of durg repurposing using genes with significant association with each phenotype when TWAS was performed based on GTEx-based model\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we conducted TWAS for 81 traits in 79,294 Korean individuals by applying both a GReX model specifically trained for Koreans based on whole blood tissue and an existing GTEx-based model. We identified 470 significant gene-trait associations (Bonferroni \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), of which 317 were novel. Pathway enrichment analysis revealed both shared and unique biological processes in the Korean-based and GTEx-based models, with the former highlighting thyroid disease and the latter emphasizing hypertension. Additionally, our computational drug repurposing analysis uncovered several promising therapeutic candidates, including clopidogrel for angina and vorinostat for fatty liver disease.\u003c/p\u003e \u003cp\u003eOur TWAS analysis utilizing a Korean-specific GReX model identified several novel gene-trait associations, highlighting potential mechanistic links based on existing literature. ANXA3, a regulator of cellular stress responses, may contribute to kidney dysfunction, potentially influencing albumin homeostasis indirectly\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. CUL7, a key regulator of skeletal growth, was associated with height, aligning with its known role in 3M syndrome\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. GPSM3, an immune signaling modulator, may affect HbA1c levels through inflammation-driven insulin resistance\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. ITIH4, an inflammation-related plasma glycoprotein, exhibited a negative correlation with inflammatory processes and coagulation, both of which are known to impact hematocrit and erythropoiesis\u003csup\u003e\u003cspan additionalcitationids=\"CR30\" citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Lastly, MT2A, a metal-binding protein, may modulate HDL-C levels by enhancing antioxidant capacity and protecting against lipid peroxidation\u003csup\u003e\u003cspan additionalcitationids=\"CR33\" citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eSubsequent pathway enrichment analysis provided further evidence of disease-related mechanisms, suggesting that certain genes within these clusters may serve as causal factors in relevant biological processes or pathways. For instance, recent evidence for thyroid disease points to possible antioxidant effects of the plant hormone auxin, which can protect thyroid membrane lipids from oxidative damage in animal models. Although auxin biosynthesis in plants is well-established, the potential connection between plant hormone and human endocrine remains underexplored. Plant hormones and their derivatives exhibit therapeutic potential, but may disrupt thyroid function if not properly managed, emphasizing the need for careful monitoring and ongoing research. Moreover, the negative regulation of DNA damage checkpoints is increasingly recognized in thyroid disease, particularly in thyroid tumors and endocrine regulation\u003csup\u003e\u003cspan additionalcitationids=\"CR36\" citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Disruptions in histidine transport have also been linked to thyroid hormone imbalances, suggesting an additional avenue for endocrine dysfunction\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eFor hypertension, the positive regulation of post-transcriptional gene silencing has identified microRNAs as crucial mediators in blood pressure regulation and cardiovascular health. MicroRNAs, which are involved in post-transcriptional gene silencing by targeting specific mRNAs for degradation or translational repression, play a significant role in hypertension, as observed in studies highlighting their regulatory impact on these biological pathways\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e,\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. For type 2 diabetes, alterations in O-glycan processing, particularly in O-GlcNAc modifications, influence insulin signaling and secretion, playing a critical role in the development of type 2 diabetes\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e,\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. Additionally, a connection between female gonad development and type 2 diabetes has been observed, as hormonal pathways in conditions like polycystic ovary syndrome (PCOS) illustrate how androgen excess can affect both reproductive and metabolic health\u003csup\u003e\u003cspan additionalcitationids=\"CR45\" citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e. The negative regulation of muscle hypertrophy, exemplified by elevated levels of myostatin, may exacerbate insulin resistance and contribute to metabolic dysfunction in type 2 diabetes\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. Furthermore, disruptions in RNA splicing processes are closely associated with insulin resistance, beta-cell dysfunction, and the onset of type 2 diabetes, offering potential targets for early diagnosis and therapeutic interventions\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. There is also evidence that fluoride exposure might impact glucose metabolism and insulin sensitivity, with effects dependent on the dose, duration, and presence of existing metabolic conditions\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e,\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. For hyperlipidemia, there is evidence that lipid imbalance can alter the immune regulation by disrupting T cell function and promoting chronic inflammation, further connecting lipid metabolism and immune homeostasis\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e,\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe results of our computational drug repurposing analysis identified several repurposed drugs with potential therapeutic applications. Using the Korean-based model, clopidogrel, a purinergic receptor antagonist, was found to reduce thrombotic events in angina by inhibiting the P2Y12 receptor and decreasing platelet aggregation\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e,\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e. Vorinostat, an HDAC inhibitor, demonstrated potential efficacy in the treatment of fatty liver disease through gene expression modulation\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e. Fulvestrant, an estrogen receptor antagonist, may influence fibroblast proliferation and inflammation, key contributors to periodontal disease progression\u003csup\u003e\u003cspan additionalcitationids=\"CR57\" citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. Rucaparib, a PARP inhibitor, emerged as a promising therapy for metastatic castration-resistant prostate cancer, particularly in BRCA-mutated cases\u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e. Pioglitazone, while lacking direct evidence for treating gastritis, may confer anti-inflammatory and antioxidant properties that alleviate gastric inflammation and mucosal damage\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e. In contrast, the GTEx-based model identified etoposide, a topoisomerase inhibitor, as potentially reducing inflammation in arthritis by modulating immune function\u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e. Enalapril, an angiotensin-converting enzyme inhibitor, showed chemopreventive properties in breast cancer by affecting angiogenesis and inflammation\u003csup\u003e\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e. Minoxidil, a KATP channel activator and vasodilator, demonstrated cardioprotective effects in heart failure\u003csup\u003e\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e,\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e. Cefotaxime, a bacterial cell wall synthesis inhibitor, may be beneficial in chronic pulmonary disease by targeting bacterial infections\u003csup\u003e\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e while azacitidine, a DNA methyltransferase inhibitor, reactivates tumor suppressor genes and enhances immune responses in colorectal cancer\u003csup\u003e\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e,\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e\u003c/sup\u003e. Simvastatin, an HMGCR inhibitor, demonstrated anti-cancer properties in lung cancer by reducing cell proliferation and metastasis\u003csup\u003e\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e,\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e\u003c/sup\u003e, and acyclovir has also shown promise in treating prostate cancer via inhibiting DNA replication\u003csup\u003e\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e,\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eBeyond known treatments like clopidogrel for angina and vorinostat for fatty liver disease, our analysis suggested novel mechanisms of existing drugs. For instance, HDAC inhibition, predominantly studied in oncology, may be effective in non-cancerous conditions such as asthma\u003csup\u003e\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e. Similarly, glucocorticoid receptors modulation, typically linked to inflammation, could be beneficial in gastric cancer, despite poorer prognoses associated with overexpression in certain tissues\u003csup\u003e\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e,\u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e\u003c/sup\u003e. Chelation therapy, particularly targeting metals like zinc and copper, demonstrates potential for managing type 2 diabetes by reducing oxidative stress and improving insulin sensitivity\u003csup\u003e\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e,\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e\u003c/sup\u003e, and it could also lower liver cancer risk\u003csup\u003e\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e,\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e\u003c/sup\u003e. Targeting tubulin polymerization has emerged as a strategy against drug-resistant tuberculosis and some cancers, including thyroid cancer, by disrupting cellular structure and division\u003csup\u003e\u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e,\u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e\u003c/sup\u003e whereas inhibition of ATPase pathways - crucial in hypoxia signaling \u0026ndash; may be applicable to thyroid cancer treatment\u003csup\u003e\u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e81\u003c/span\u003e\u003c/sup\u003e. Acetylcholine receptor antagonists could regulate pain and inflammation in cystitis models, suggesting its therapeutic potential in bladder conditions\u003csup\u003e\u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e,\u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e83\u003c/span\u003e\u003c/sup\u003e. Furthermore, mTOR inhibition, particularly mTORC1, shows potential in treating fatty liver disease by reducing lipid accumulation and enhancing fatty acid metabolism\u003csup\u003e\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e,\u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e85\u003c/span\u003e\u003c/sup\u003e. Although this study focused on the top candidate drug for each phenotype, future research should explore combination therapies and alternative drugs with slightly lower connectivity scores to uncover synergistic benefits or novel therapeutic strategies. Further experimental validation will be necessary to confirm the therapeutic potential of these repurposed drugs in various disease contexts.\u003c/p\u003e \u003cp\u003eDespite these promising findings, our study has certain limitations. First, our reliance on whole blood tissue as the sole source for the GReX model may skew results for traits less relevant to this tissue, such as thyroid cancer. Including multi-tissue or cell-type specific data would address this limitation\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Second, we constructed our Korean GReX model using imputed SNP chip data rather than whole-genome sequencing. Although we applied stringent quality control measures (imputation info score\u0026thinsp;\u0026gt;\u0026thinsp;0.8), imputation-related errors and biases in effect size estimates remain potential challenges\u003csup\u003e\u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e86\u003c/span\u003e\u003c/sup\u003e. Finally, the lack of extensive replication analyses for novel signals limits the generalizability of our findings. Further replication involving larger East Asian cohorts are necessary to validate these associations.\u003c/p\u003e \u003cp\u003eIn conclusion, our large-scale TWAS on a Korean population offers important insights into the genetic architecture of complex traits, uncovering multiple novel gene-trait associations with implications for disease prevention, diagnosis, and treatment. The pathway-level analyses underscore the value of ancestry-specific models, which can reveal previously unrecognized mechanisms and deepen our understanding of how genetic factors contribute to complex diseases. By integrating computational approaches, we have also identified opportunities to repurpose existing drugs, potentially expediting the translation of genetic discoveries into clinical applications. Ultimately, by focusing on underrepresented population, our research underscores the importance of incorporating diverse genetic data to enhance the accuracy, relevance, and inclusivity of genetic studies. This study not only advances our understanding of population-specific genetic architectures but also lays the foundation for globally impactful discoveries that can inform precision medicine and improve health outcomes for diverse populations.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eGReX Training Dataset Information\u003c/h2\u003e \u003cp\u003eWe combined two datasets for Korean GReX model training. The first dataset included genotypes and whole blood RNA sequencing data from 436 Korean subjects with mild asthma symptoms like dyspnea, wheezing, or cough for over three months, and airway hyperresponsiveness but without severe lung damage, bronchiectasis, or a history of lung resection\u003csup\u003e\u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e87\u003c/span\u003e\u003c/sup\u003e. The second dataset was drawn from participants in the Gwangju Alzheimer\u0026rsquo;s \u0026amp; Related Dementia (GARD) cohort registry at Chosun University in Gwangju, Korea, and included genotypes and RNA sequencing data from 300 individuals with mild cognitive impairment \u003csup\u003e\u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e88\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eGenotype and RNA Quality Control in GReX Training Dataset\u003c/h2\u003e \u003cp\u003eAsan and Chosun study participants were genotyped using the Korea Biobank Array, which was designed by the Korea National Institute of Health (KNIH)\u003csup\u003e\u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e\u003c/sup\u003e. Both datasets were preprocessed as follows. A K-medoid clustering algorithm was employed to identify the genotypes, minimizing batch effects\u003csup\u003e\u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e90\u003c/span\u003e\u003c/sup\u003e. Quality control procedures were executed using PLINK and ONETOOL\u003csup\u003e\u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e91\u003c/span\u003e\u003c/sup\u003e, with exclusion criteria set for samples with sex discrepancies or call rates below 90%. SNPs were filtered based on a call rate threshold of less than 95% or deviation from the Hardy\u0026ndash;Weinberg equilibrium (HWE) test (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;1 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e). Imputation was conducted with IMPUTE2\u003csup\u003e92\u003c/sup\u003e using the NARD V1 imputation server\u003csup\u003e\u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e93\u003c/span\u003e\u003c/sup\u003e. Pre-phasing was performed with Eagle v2.4.1\u003csup\u003e94\u003c/sup\u003e, and imputation was done with Minimac4\u003csup\u003e95\u003c/sup\u003e, both using default settings. Post-imputation quality control was used to exclude imputed SNPs with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}\\)\u003c/span\u003e\u003c/span\u003e \u0026lt; 0.8, duplicates, missing genotype rates \u0026gt; 0.05, HWE P \u0026lt; 1 \u0026times; 10\u003csup\u003e\u0026minus;6\u003c/sup\u003e, and minor allele frequencies (MAF) \u0026lt;0.01. After QC, SNP coordinates were updated to the hg38 assembly using the R package \u0026ldquo;rtracklayer\u0026rdquo; (version 1.46.0). Then, the two data sets were merged and SNPs were excluded with missing genotype rates \u0026gt; 0.05, HWE P \u0026lt; 1 \u0026times; 10\u003csup\u003e\u0026minus;6\u003c/sup\u003e, and MAF \u0026lt;\u0026thinsp;0.01. Finally, 4,535,155 SNPs for 674 samples were used for GReX modeling.\u003c/p\u003e \u003cp\u003eRNA sequencing data were also available for these Asan and Chosun study participants, with detailed descriptions of RNA library preparation and sequencing provided in the previous publications\u003csup\u003e\u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e87\u003c/span\u003e,\u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e88\u003c/span\u003e\u003c/sup\u003e. Briefly, transcriptomic QC involved trimming low-quality reads and adapter sequences with Trimmomatic v0.39\u003csup\u003e96\u003c/sup\u003e. The filtered reads were aligned to the GRCh38/hg38 human reference genome using STAR v2.7.8a \u003csup\u003e\u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e97\u003c/span\u003e\u003c/sup\u003e, and gene-level expression was quantified using RNA-SeQC v2.4.2\u003csup\u003e98\u003c/sup\u003e with GENCODE Release 26 annotation (gencode.v26.GRCh38.genes.gtf)\u003csup\u003e\u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e99\u003c/span\u003e\u003c/sup\u003e. Genes with a Transcripts per Million (TPM) value\u0026thinsp;\u0026lt;\u0026thinsp;0.1 in more than 80% of the participants were excluded. The resulting gene expression profiles were normalized by TMM\u003csup\u003e\u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e100\u003c/span\u003e\u003c/sup\u003e, followed by inverse normal transformation. To correct for batch effects and experimental variability, PEER\u003csup\u003e\u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e101\u003c/span\u003e\u003c/sup\u003e factor analysis was conducted. Finally, normalized gene expression values were adjusted for sex, the top 3 genotype principal components(PCs), and the top 60 PEER factors, as recommended by the GTEx Consortium\u003csup\u003e\u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e102\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eGenotype Quality Control for TWAS Study Samples\u003c/h2\u003e \u003cp\u003eWe conducted TWAS on the KoGES and GENIE cohorts, both of which were genotyped using the Korea Biobank Array designed by the Korea National Institute of Health (KNIH)\u003csup\u003e\u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e\u003c/sup\u003e. Variant calling, quality control (QC), and imputation procedures followed the same protocol used for the GReX training dataset. After merging both cohorts, 3,504,297 SNPs across 81,153 subjects remained for analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eBuilding GReX Prediction Models\u003c/h2\u003e \u003cp\u003eWe constructed the predictive models for GReX by integrating genotype and transcriptome data (Asan\u0026thinsp;+\u0026thinsp;Chosun) and focusing on cis-region SNPs (\u0026plusmn;\u0026thinsp;1 Mb of a gene\u0026rsquo;s start or end position) for downstream TWAS analysis. Model development was conducted using the PredictDB pipeline, which automates data preprocessing, training, and integration of the elastic net regression results into database files used as weights for predicting gene expression\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e This pipeline utilizes the R package \u0026ldquo;glmnet\u0026rdquo; (version 4.1.2), designed to model gene expression predictions based on cis-region SNPs\u003csup\u003e\u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e103\u003c/span\u003e\u003c/sup\u003e. The elastic net regularization is controlled by the shrinkage parameter alpha (α), where α\u0026thinsp;=\u0026thinsp;0 corresponds to ridge regression and α\u0026thinsp;=\u0026thinsp;1 corresponds to lasso regression. In our model, α was set to 0.5. Sex, age, 3 genotype-based PCs, and 60 PEER factors were included as covariates to account for possible confounding.\u003c/p\u003e \u003cp\u003eTo evaluate the predictive accuracy, we employed a nested cross-validation. Our dataset was partitioned into five equally sized folds. For each fold, we applied a 10-fold cross-validated elastic net model to 80% of the data, simultaneously tuning the lambda parameter. We then assessed model performance predicted on the remaining 20% of the data by comparing predicted and observed gene expression to compute the coefficient of determination (\u0026#119877;\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e). We repeated this process for all five folds, ultimately deriving the average \u0026#119877;\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, the Pearson correlation coefficient, and a combined Z score via Stouffer\u0026rsquo;s method as metrics of overall model performance across the five folds.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eCis-eQTL analysis\u003c/h2\u003e \u003cp\u003eWe performed a cis-eQTL analysis in the previously described GReX training dataset. Cis-region SNPs were defined as those residing within \u0026plusmn;\u0026thinsp;1 Mb of a gene\u0026rsquo;s start or end position, according to the GENCODE Release 26 (hg38). To identify cis-eQTLs, we employed the Matrix eQTL R package (version 3.2)\u003csup\u003e\u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e104\u003c/span\u003e\u003c/sup\u003e with default settings, which applies a linear regression framework to test SNP-gene expression associations. In this analysis, gene expression values were adjusted for sex, age, 3 genotype-based PCs, and 60 PEER factors to account for confounding factors.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eCis-heritability calculation of gene expression\u003c/h2\u003e \u003cp\u003eWe constructed a genetic relationship matrix for each gene using SNPs with MAF\u0026thinsp;\u0026gt;\u0026thinsp;0.05, located within \u0026plusmn;\u0026thinsp;1 Mb of the gene and meeting HWE (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;1 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e). From these genetic relationship matrixes, we applied a variance-component model to estimate narrow-sense heritability (ℎ\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e) of gene expression for each gene. This analysis was performed separately for both the GTEx-based model and the Korean (Asan\u0026thinsp;+\u0026thinsp;Chosun) model. We used the restricted maximum likelihood (REML) approach as implemented in the GCTA\u003csup\u003e\u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e105\u003c/span\u003e\u003c/sup\u003e software (GCTA-GREML). For the heritability calculations, gene expression values were adjusted for sex, age, 3 genotype-based PCs, and 60 PEER factors.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eTranscriptome-Wide Association Study (TWAS) of 81 Traits\u003c/h2\u003e \u003cp\u003eFor association analysis, we retained only those genes with a cross validated \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}\u0026gt;0.05\\)\u003c/span\u003e\u003c/span\u003e in each model, yielding 1,146 genes for GTEx and 1,678 genes for Korean-based model. Quantitative traits were inverse rank-normalized using the R package \u0026ldquo;RNOmni\u0026rdquo; (version 1.0.0), except for anthropometric and lifestyle traits. We performed TWAS on 30 quantitative traits using linear regression and on 51 binary traits using Firth\u0026rsquo;s logistic regression through the R package \u0026ldquo;logistf\u0026rdquo; (version 1.24.1) to mitigate type I error due to case-control imbalance. Multiple testing was addressed by applying separate Bonferroni-corrected significance thresholds for each model: \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;4.36\u0026times;10\u003csup\u003e\u0026minus;5\u003c/sup\u003e (0.05/1,146) for the GTEx-based model and \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;2.97\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e (0.05/1,678) for the Korean-based model. All regression models were adjusted for age, sex, body mass index (BMI), smoking status, and the top 10 PCs.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eIdentification of Novel Gene-Trait Associations and Replication\u003c/h2\u003e \u003cp\u003eWe used Harmonizome\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e to assess whether any of the gene-trait associations identified in our TWAS results had been previously reported. Using Harmonizome, we categorized gene-trait associations into four categories: Complete, Strong, Weak, and No (Table S6). This categorization enabled us to systematically identify and propose novel gene-trait associations.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003ePathway Enrichment Analysis\u003c/h2\u003e \u003cp\u003eWe performed pathway enrichment analysis to evaluate the biological relevance of genes identified through TWAS for each disease (FDR \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.1). As the primary resource for enrichment, we first considered the GO\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e gene set collection from the C5 category in MSigDB\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Only traits associated with at least three genes in gene-trait association tests were included in the analysis. We used a one-sided Fisher\u0026rsquo;s exact test to determine the statistical significance of GO terms, with \u003cem\u003eP\u003c/em\u003e adjusted via Benjamini-Hochberg method (FDR \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.1). Next, we employed REVIGO\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e to reduce redundancy among enriched GO terms, and the refined results were visualized using CirGO\u003csup\u003e\u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e106\u003c/span\u003e\u003c/sup\u003e for a more intuitive and comprehensive presentation. REVIGO applies a clustering algorithm based on semantic similarity drawing upon pre-computed information content to streamline the interpretation and visualization of GO analysis outcomes. This approach selects a representative subset of terms, facilitating clearer insights into the underlying biological processes. These enrichment analyses were repeated for KEGG\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, Reactome\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e, and WikiPathways\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, focusing on canonical pathways from the C2 category in MSigDB\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eComputational Drug Repurposing Analysis\u003c/h2\u003e \u003cp\u003eWe conducted a computational drug repurposing analysis using the CMap approach\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, a web-based tool for drug repositioning that matches input gene signatures (both up- and down-regulated) with drug-induced gene expression signatures derived from in vitro experiments (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://clue.io/\u003c/span\u003e\u003cspan address=\"https://clue.io/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). For each trait, we selected the top 50 up-regulated and bottom 50 down-regulated genes identified from the TWAS for drug repurposing analysis. Connectivity scores were calculated using Kolmogorov-Smirnov statistics, measuring the association between the TWAS-derived gene expression profile for a given phenotype and the corresponding drug-induced gene expression profile in the CMap database. To ensure clinical relevance, we restricted our analyses to FDA-approved drugs using DrugBank (version 5.1.12, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://go.drugbank.com\u003c/span\u003e\u003cspan address=\"https://go.drugbank.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e)\u003csup\u003e107\u003c/sup\u003e, and obtained their mechanisms of action from the CLUE Drug Repurposing Hub\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Ultimately, we identified compounds whose gene expression profiles opposed the disease-risk genes detected through TWAS.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eEthical approval and informed consent\u003c/b\u003e \u003c/p\u003e \u003c/p\u003e \u003cp\u003e This study was approved by Institutional Review Board of Seoul National University (E2410/001\u0026ndash;004). All participants or their legal guardians provided written informed consent at the time of cohort enrollment. The study using data from Asan Medical Center was approved by the Institutional Review Board (IRB) of Asan Medical Center (2019\u0026thinsp;\u0026minus;\u0026thinsp;0376), and the study using data from Chosun University Hospital, Korea (CHOSUN 2013\u0026ndash;12\u0026ndash;018\u0026ndash;070). For further details, please refer to previous studies\u003csup\u003e\u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e87\u003c/span\u003e,\u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e88\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e "},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIndividual-level phenotypic and genetic data for KOGES (https://is.kdca.go.kr/) are available upon application. PrediXcan Elastic-Net base models trained on GTEx V8 can be accessed from the PredictDB Data Repository (https://predictdb.org/). The weights for the Korean GReX model and summary statistics for all TWAS analyses are available upon request.\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll code and computational tools used in this study are publicly available.\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMH, SK, and SW conceptualized and designed the study. MH, YJ, AD, TK, JG, KL, SK and SW contributed to data acquisition and analysis. MH, HJ, KP, NK, JA, SK, SW carried out the statistical analysis and interpretation. MH, SK, and SW wrote and revised the manuscript. All authors have provided critical review and edit for the manuscript. \u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government(MSIT)(RS-2024-00346850) and by the National Research Foundation (NRF) grant (NRF-2021R1A5A1033157) funded by the Korean government (MSIT). This work was conducted with bioresources from National Biobank of Korea, the Korea Disease Control and Prevention Agency, Republic of Korea (NBK-2020-101). This work was supported by the national supercomputing center with supercomputing resources including technical support (KSC-2023-CHA-0033), and large data transfer was supported by KREONET, which is managed and operated by KISTI. Min Heo was supported by the Hyundai Motor Chung Mong-Koo Foundation.\u003c/p\u003e\n\n"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eGallagher, M. D. \u0026amp; Chen-Plotkin, A. S. The Post-GWAS Era: From Association to Function. \u003cem\u003eAm J Hum Genet\u003c/em\u003e \u003cstrong\u003e102\u003c/strong\u003e, 717\u0026ndash;730 (2018).\u003c/li\u003e\n\u003cli\u003eWainberg, M. \u003cem\u003eet al.\u003c/em\u003e Opportunities and challenges for transcriptome-wide association studies. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e51\u003c/strong\u003e, 592\u0026ndash;599 (2019).\u003c/li\u003e\n\u003cli\u003eGamazon, E. R. \u003cem\u003eet al.\u003c/em\u003e A gene-based association method for mapping traits using reference transcriptome data. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, 1091\u0026ndash;1098 (2015).\u003c/li\u003e\n\u003cli\u003eGusev, A. \u003cem\u003eet al.\u003c/em\u003e Integrative approaches for large-scale transcriptome-wide association studies. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 245\u0026ndash;252 (2016).\u003c/li\u003e\n\u003cli\u003eIshigaki, K. \u003cem\u003eet al.\u003c/em\u003e Large-scale genome-wide association study in a Japanese population identifies novel susceptibility loci across different diseases. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, 669\u0026ndash;679 (2020).\u003c/li\u003e\n\u003cli\u003eMai, J., Lu, M., Gao, Q., Zeng, J. \u0026amp; Xiao, J. Transcriptome-wide association studies: recent advances in methods, applications and available databases. \u003cem\u003eCommun Biol\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 1\u0026ndash;10 (2023).\u003c/li\u003e\n\u003cli\u003eWojcik, G. L. \u003cem\u003eet al.\u003c/em\u003e Genetic analyses of diverse populations improves discovery for complex traits. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e570\u003c/strong\u003e, 514\u0026ndash;518 (2019).\u003c/li\u003e\n\u003cli\u003eTraylor, M. \u0026amp; Lewis, C. M. Genetic discovery in multi-ethnic populations. \u003cem\u003eEuropean Journal of Human Genetics\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 1097\u0026ndash;1098 (2016).\u003c/li\u003e\n\u003cli\u003eMogil, L. S. \u003cem\u003eet al.\u003c/em\u003e Genetic architecture of gene expression traits across diverse populations. \u003cem\u003ePLOS Genetics\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, e1007586 (2018).\u003c/li\u003e\n\u003cli\u003eKim, Y., Han, B.-G., \u0026amp; the KoGES group. Cohort Profile: The Korean Genome and Epidemiology Study (KoGES) Consortium. \u003cem\u003eInternational Journal of Epidemiology\u003c/em\u003e \u003cstrong\u003e46\u003c/strong\u003e, 1350 (2017).\u003c/li\u003e\n\u003cli\u003eLee, C. \u003cem\u003eet al.\u003c/em\u003e Health and Prevention Enhancement (H-PEACE): a retrospective, population-based cohort study conducted at the Seoul National University Hospital Gangnam Center, Korea. \u003cem\u003eBMJ Open\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, e019327 (2018).\u003c/li\u003e\n\u003cli\u003eRouillard, A. D. \u003cem\u003eet al.\u003c/em\u003e The harmonizome: a collection of processed datasets gathered to serve and mine knowledge about genes and proteins. \u003cem\u003eDatabase\u003c/em\u003e \u003cstrong\u003e2016\u003c/strong\u003e, baw100 (2016).\u003c/li\u003e\n\u003cli\u003eLiberzon, A. \u003cem\u003eet al.\u003c/em\u003e The Molecular Signatures Database Hallmark Gene Set Collection. \u003cem\u003eCell Systems\u003c/em\u003e \u003cstrong\u003e1\u003c/strong\u003e, 417\u0026ndash;425 (2015).\u003c/li\u003e\n\u003cli\u003eSubramanian, A. \u003cem\u003eet al.\u003c/em\u003e A Next Generation Connectivity Map: L1000 Platform and the First 1,000,000 Profiles. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e171\u003c/strong\u003e, 1437-1452.e17 (2017).\u003c/li\u003e\n\u003cli\u003eCorsello, S. M. \u003cem\u003eet al.\u003c/em\u003e The Drug Repurposing Hub: a next-generation drug library and information resource. \u003cem\u003eNat Med\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 405\u0026ndash;408 (2017).\u003c/li\u003e\n\u003cli\u003eKarczewski, K. J. \u003cem\u003eet al.\u003c/em\u003e The mutational constraint spectrum quantified from variation in 141,456 humans. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e581\u003c/strong\u003e, 434\u0026ndash;443 (2020).\u003c/li\u003e\n\u003cli\u003eKarczewski, K. J. \u003cem\u003eet al.\u003c/em\u003e The ExAC browser: displaying reference data information from over 60 000 exomes. \u003cem\u003eNucleic Acids Research\u003c/em\u003e \u003cstrong\u003e45\u003c/strong\u003e, D840\u0026ndash;D845 (2017).\u003c/li\u003e\n\u003cli\u003eSupek, F., Bo\u0026scaron;njak, M., \u0026Scaron;kunca, N. \u0026amp; \u0026Scaron;muc, T. REVIGO Summarizes and Visualizes Long Lists of Gene Ontology Terms. \u003cem\u003ePLOS ONE\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, e21800 (2011).\u003c/li\u003e\n\u003cli\u003eAshburner, M. \u003cem\u003eet al.\u003c/em\u003e Gene Ontology: tool for the unification of biology. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 25\u0026ndash;29 (2000).\u003c/li\u003e\n\u003cli\u003eKanehisa, M. \u0026amp; Goto, S. KEGG: Kyoto Encyclopedia of Genes and Genomes. \u003cem\u003eNucleic Acids Research\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 27\u0026ndash;30 (2000).\u003c/li\u003e\n\u003cli\u003eMilacic, M. \u003cem\u003eet al.\u003c/em\u003e The Reactome Pathway Knowledgebase 2024. \u003cem\u003eNucleic Acids Research\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, D672\u0026ndash;D678 (2024).\u003c/li\u003e\n\u003cli\u003eAgrawal, A. \u003cem\u003eet al.\u003c/em\u003e WikiPathways 2024: next generation pathway database. \u003cem\u003eNucleic Acids Research\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, D679\u0026ndash;D689 (2024).\u003c/li\u003e\n\u003cli\u003eLambers Heerspink, H. J. \u0026amp; Gansevoort, R. T. Albuminuria Is an Appropriate Therapeutic Target in Patients with CKD: The Pro View. \u003cem\u003eClinical Journal of the American Society of Nephrology\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 1079 (2015).\u003c/li\u003e\n\u003cli\u003eZhao, J., Chang, L., Tu, J., Sun, B. \u0026amp; Wei, X. Evaluation of Annexins Family as Potential Biomarker for Predicting Progression and Prognosis in Clear Renal Cell Carcinoma. \u003cem\u003eEvidence-Based Complementary and Alternative Medicine\u003c/em\u003e \u003cstrong\u003e2022\u003c/strong\u003e, 8748434 (2022).\u003c/li\u003e\n\u003cli\u003eShaikh, S. \u003cem\u003eet al.\u003c/em\u003e Novel mutation in Cul7 gene in a family diagnosed with 3M syndrome. \u003cem\u003eJ Genet\u003c/em\u003e \u003cstrong\u003e98\u003c/strong\u003e, 21 (2019).\u003c/li\u003e\n\u003cli\u003eDeeb, A., Afandi, O., Attia, S. \u0026amp; Fatih, A. E. 3-M syndrome: a novel CUL7 mutation associated with respiratory distress and a good response to GH therapy. \u003cem\u003eEndocrinol Diabetes Metab Case Rep\u003c/em\u003e \u003cstrong\u003e2015\u003c/strong\u003e, 150012 (2015).\u003c/li\u003e\n\u003cli\u003eSlieker, R. C. \u003cem\u003eet al.\u003c/em\u003e HbA1c is associated with altered expression in blood of cell cycle- and immune response-related genes. \u003cem\u003eDiabetologia\u003c/em\u003e \u003cstrong\u003e61\u003c/strong\u003e, 138\u0026ndash;146 (2018).\u003c/li\u003e\n\u003cli\u003eFan, R. \u003cem\u003eet al.\u003c/em\u003e Loss of the co-repressor GPS2 sensitizes macrophage activation upon metabolic stress induced by obesity and type 2 diabetes. \u003cem\u003eNat Med\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 780\u0026ndash;791 (2016).\u003c/li\u003e\n\u003cli\u003eHuo, Y., Lai, Y., Feng, Q., Wang, Q. \u0026amp; Li, J. Serum ITIH4 in coronary heart disease: a potential anti-inflammatory biomarker related to stenosis degree and risk of major adverse cardiovascular events. \u003cem\u003eBiomark Med\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 1279\u0026ndash;1288 (2022).\u003c/li\u003e\n\u003cli\u003eMa, Y. \u003cem\u003eet al.\u003c/em\u003e ITIH4, as an inflammation biomarker, mainly increases in bacterial bloodstream infection. \u003cem\u003eCytokine\u003c/em\u003e \u003cstrong\u003e138\u003c/strong\u003e, 155377 (2021).\u003c/li\u003e\n\u003cli\u003eTian, L., Zhao, S. \u0026amp; Zhang, R. ITIH4 is a predictor for coronary thrombus in coronary arteriography patients. \u003cem\u003eFuture Cardiol\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 547\u0026ndash;554 (2024).\u003c/li\u003e\n\u003cli\u003eLing, X.-B. \u003cem\u003eet al.\u003c/em\u003e Mammalian Metallothionein-2A and Oxidative Stress. \u003cem\u003eInt J Mol Sci\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 1483 (2016).\u003c/li\u003e\n\u003cli\u003eŚciskalska, M., Ołdakowska, M. \u0026amp; Milnerowicz, H. Importance of Genetic Polymorphisms in MT1 and MT2 Genes in Metals Homeostasis and Their Relationship with the Risk of Acute Pancreatitis Occurrence in Smokers-Preliminary Findings. \u003cem\u003eInt J Mol Sci\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 5725 (2021).\u003c/li\u003e\n\u003cli\u003eYang, C.-C. \u003cem\u003eet al.\u003c/em\u003e The association of the blood lead level and serum lipid concentrations may be modified by the genetic combination of the metallothionein 2A polymorphisms rs10636 GC and rs28366003 AA. \u003cem\u003eJ Clin Lipidol\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 234\u0026ndash;241 (2017).\u003c/li\u003e\n\u003cli\u003eKarger, S. \u003cem\u003eet al.\u003c/em\u003e Distinct pattern of oxidative DNA damage and DNA repair in follicular thyroid tumours. \u003cem\u003eJ Mol Endocrinol\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 193\u0026ndash;202 (2012).\u003c/li\u003e\n\u003cli\u003eKim, H.-J. \u0026amp; Yun, J. Checkpoint-dependent phosphorylation of Med1/TRAP220 in response to DNA damage. \u003cem\u003eActa Biochim Biophys Sin (Shanghai)\u003c/em\u003e \u003cstrong\u003e49\u003c/strong\u003e, 496\u0026ndash;502 (2017).\u003c/li\u003e\n\u003cli\u003eSperka, T., Wang, J. \u0026amp; Rudolph, K. L. DNA damage checkpoints in stem cells, ageing and cancer. \u003cem\u003eNat Rev Mol Cell Biol\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 579\u0026ndash;590 (2012).\u003c/li\u003e\n\u003cli\u003eBraun, D., Lelios, I., Krause, G. \u0026amp; Schweizer, U. Histidines in potential substrate recognition sites affect thyroid hormone transport by monocarboxylate transporter 8 (MCT8). \u003cem\u003eEndocrinology\u003c/em\u003e \u003cstrong\u003e154\u003c/strong\u003e, 2553\u0026ndash;2561 (2013).\u003c/li\u003e\n\u003cli\u003eFiletti, S. \u0026amp; Rapoport, B. Inhibitors of specific aminoacyl-tRNA synthetases prevent thyrotropin-induced desensitization in cultured human thyroid cells. \u003cem\u003eJ Biol Chem\u003c/em\u003e \u003cstrong\u003e257\u003c/strong\u003e, 1342\u0026ndash;1346 (1982).\u003c/li\u003e\n\u003cli\u003eKontaraki, J. E., Marketou, M. E., Zacharis, E. A., Parthenakis, F. I. \u0026amp; Vardas, P. E. MicroRNA-9 and microRNA-126 expression levels in patients with essential hypertension: potential markers of target-organ damage. \u003cem\u003eJ Am Soc Hypertens\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 368\u0026ndash;375 (2014).\u003c/li\u003e\n\u003cli\u003eHuang, Y. \u003cem\u003eet al.\u003c/em\u003e Circulating miR-92a expression level in patients with essential hypertension: a potential marker of atherosclerosis. \u003cem\u003eJ Hum Hypertens\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 200\u0026ndash;205 (2017).\u003c/li\u003e\n\u003cli\u003eKaleem, A. \u003cem\u003eet al.\u003c/em\u003e Phosphorylated and O-GlcNAc Modified IRS-1 (Ser1101) and -2 (Ser1149) Contribute to Human Diabetes Type II. \u003cem\u003eProtein Pept Lett\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 333\u0026ndash;339 (2021).\u003c/li\u003e\n\u003cli\u003eOhtsubo, K. \u003cem\u003eet al.\u003c/em\u003e N-Glycosylation modulates the membrane sub-domain distribution and activity of glucose transporter 2 in pancreatic beta cells. \u003cem\u003eBiochem Biophys Res Commun\u003c/em\u003e \u003cstrong\u003e434\u003c/strong\u003e, 346\u0026ndash;351 (2013).\u003c/li\u003e\n\u003cli\u003eGambineri, A. \u0026amp; Pelusi, C. Sex hormones, obesity and type 2 diabetes: is there a link? \u003cem\u003eEndocr Connect\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, R1\u0026ndash;R9 (2019).\u003c/li\u003e\n\u003cli\u003eNavarro, G. \u003cem\u003eet al.\u003c/em\u003e Androgen excess in pancreatic \u0026beta; cells and neurons predisposes female mice to type 2 diabetes. \u003cem\u003eJCI Insight\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, e98607, 98607 (2018).\u003c/li\u003e\n\u003cli\u003eLubura, M. \u003cem\u003eet al.\u003c/em\u003e Diabetes prevalence in NZO females depends on estrogen action on liver fat content. \u003cem\u003eAm J Physiol Endocrinol Metab\u003c/em\u003e \u003cstrong\u003e309\u003c/strong\u003e, E968-980 (2015).\u003c/li\u003e\n\u003cli\u003eBrandt, C. \u003cem\u003eet al.\u003c/em\u003e Plasma and muscle myostatin in relation to type 2 diabetes. \u003cem\u003ePLoS One\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, e37236 (2012).\u003c/li\u003e\n\u003cli\u003eGahete, M. D. \u003cem\u003eet al.\u003c/em\u003e Changes in Splicing Machinery Components Influence, Precede, and Early Predict the Development of Type 2 Diabetes: From the CORDIOPREV Study. \u003cem\u003eEBioMedicine\u003c/em\u003e \u003cstrong\u003e37\u003c/strong\u003e, 356\u0026ndash;365 (2018).\u003c/li\u003e\n\u003cli\u003eKumar, P., Gupta, R. \u0026amp; Gupta, A. Vitamin D deficiency in patients with diabetes and its correlation with water fluoride levels. \u003cem\u003eJ Water Health\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 125\u0026ndash;137 (2023).\u003c/li\u003e\n\u003cli\u003eLobo, J. G. V. M. \u003cem\u003eet al.\u003c/em\u003e Low-Level Fluoride Exposure Increases Insulin Sensitivity in Experimental Diabetes. \u003cem\u003eJ Dent Res\u003c/em\u003e \u003cstrong\u003e94\u003c/strong\u003e, 990\u0026ndash;997 (2015).\u003c/li\u003e\n\u003cli\u003eBagley, J., Yuan, J., Chandrakar, A. \u0026amp; Iacomini, J. Hyperlipidemia Alters Regulatory T Cell Function and Promotes Resistance to Tolerance Induction Through Costimulatory Molecule Blockade. \u003cem\u003eAm J Transplant\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 2324\u0026ndash;2335 (2015).\u003c/li\u003e\n\u003cli\u003eChen, S. \u003cem\u003eet al.\u003c/em\u003e Hyperlipidemia modifies innate immune responses to lipopolysaccharide via the TLR-NF-\u0026kappa;B signaling pathway. \u003cem\u003eInflammation\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 968\u0026ndash;976 (2013).\u003c/li\u003e\n\u003cli\u003eMehta, S. R., Yusuf, S., \u0026amp; Clopidogrel in Unstable angina to prevent Recurrent Events (CURE) Study Investigators. The Clopidogrel in Unstable angina to prevent Recurrent Events (CURE) trial programme; rationale, design and baseline characteristics including a meta-analysis of the effects of thienopyridines in vascular disease. \u003cem\u003eEur Heart J\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 2033\u0026ndash;2041 (2000).\u003c/li\u003e\n\u003cli\u003eSavi, P. \u0026amp; Herbert, J.-M. Clopidogrel and Ticlopidine: P2Y\u003csub\u003e12\u003c/sub\u003e Adenosine Diphosphate-Receptor Antagonists for the Prevention of Atherothrombosis. \u003cem\u003eSemin Thromb Hemost\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 174\u0026ndash;183 (2005).\u003c/li\u003e\n\u003cli\u003eCheng, Z. \u003cem\u003eet al.\u003c/em\u003e Gene expression profile-based drug screen identifies SAHA as a novel treatment for NAFLD. \u003cem\u003eMol. Omics\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 50\u0026ndash;58 (2019).\u003c/li\u003e\n\u003cli\u003eSteffens, J. P. \u003cem\u003eet al.\u003c/em\u003e Role of testosterone and androgen receptor in periodontal disease progression in female rats. \u003cem\u003eJournal of Periodontology\u003c/em\u003e \u003cstrong\u003e91\u003c/strong\u003e, 545\u0026ndash;553 (2020).\u003c/li\u003e\n\u003cli\u003eShapiro, L. F. \u0026amp; Freeman, K. The relationship between estrogen, estrogen receptors and periodontal disease in adult women. \u003cem\u003eJ Mich Dent Assoc\u003c/em\u003e \u003cstrong\u003e96\u003c/strong\u003e, 40\u0026ndash;44 (2014).\u003c/li\u003e\n\u003cli\u003eSantana, L. C. L. \u003cem\u003eet al.\u003c/em\u003e Testosterone Increases Fibroblast Proliferation in vitro Through Androgen and Estrogen Receptor Activation. \u003cem\u003eJ Int Acad Periodontol\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 146\u0026ndash;155 (2020).\u003c/li\u003e\n\u003cli\u003eFizazi, K. \u003cem\u003eet al.\u003c/em\u003e Rucaparib or Physician\u0026rsquo;s Choice in Metastatic Prostate Cancer. \u003cem\u003eN Engl J Med\u003c/em\u003e \u003cstrong\u003e388\u003c/strong\u003e, 719\u0026ndash;732 (2023).\u003c/li\u003e\n\u003cli\u003eSaha, L. Role of peroxisome proliferator-activated receptors alpha and gamma in gastric ulcer: An overview of experimental evidences. \u003cem\u003eWJGPT\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 120 (2015).\u003c/li\u003e\n\u003cli\u003eVerdrengh, M., Isaksson, O. \u0026amp; Tarkowski, A. Topoisomerase II inhibitors, irrespective of their chemical composition, ameliorate experimental arthritis. \u003cem\u003eRheumatology\u003c/em\u003e \u003cstrong\u003e44\u003c/strong\u003e, 183\u0026ndash;186 (2005).\u003c/li\u003e\n\u003cli\u003eNapoleone, E. \u003cem\u003eet al.\u003c/em\u003e Inhibition of the renin-angiotensin system downregulates tissue factor and vascular endothelial growth factor in human breast carcinoma cells. \u003cem\u003eThrombosis Research\u003c/em\u003e \u003cstrong\u003e129\u003c/strong\u003e, 736\u0026ndash;742 (2012).\u003c/li\u003e\n\u003cli\u003eSato, T., Li, Y., Saito, T. \u0026amp; Nakaya, H. Minoxidil opens mitochondrial KATP channels and confers cardioprotection. \u003cem\u003eBritish Journal of Pharmacology\u003c/em\u003e \u003cstrong\u003e141\u003c/strong\u003e, 360\u0026ndash;366 (2004).\u003c/li\u003e\n\u003cli\u003eNichols, C. G., Singh, G. K. \u0026amp; Grange, D. K. KATP Channels and Cardiovascular Disease. \u003cem\u003eCirculation Research\u003c/em\u003e \u003cstrong\u003e112\u003c/strong\u003e, 1059\u0026ndash;1072 (2013).\u003c/li\u003e\n\u003cli\u003eVan Zanten, A. R. H. \u003cem\u003eet al.\u003c/em\u003e Continuous \u003cem\u003evs.\u003c/em\u003e intermittent cefotaxime administration in patients with chronic obstructive pulmonary disease and respiratory tract infections: pharmacokinetics/pharmacodynamics, bacterial susceptibility and clinical efficacy. \u003cem\u003eBrit J Clinical Pharma\u003c/em\u003e \u003cstrong\u003e63\u003c/strong\u003e, 100\u0026ndash;109 (2007).\u003c/li\u003e\n\u003cli\u003eLi, H. \u003cem\u003eet al.\u003c/em\u003e Immune regulation by low doses of the DNA methyltransferase inhibitor 5-azacitidine in common human epithelial cancers. \u003cem\u003eOncotarget\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 587\u0026ndash;598 (2014).\u003c/li\u003e\n\u003cli\u003eHsi, L. C., Xi, X., Wu, Y. \u0026amp; Lippman, S. M. The methyltransferase inhibitor 5-aza-2-deoxycytidine induces apoptosis via induction of 15-lipoxygenase-1 in colorectal cancer cells. \u003cem\u003eMolecular Cancer Therapeutics\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 1740\u0026ndash;1746 (2005).\u003c/li\u003e\n\u003cli\u003eLi, Y., Fu, J., Yuan, X. \u0026amp; Hu, C. Simvastatin inhibits the proliferation of A549 lung cancer cells through oxidative stress and up-regulation of SOD2. \u003cem\u003ePharmazie\u003c/em\u003e \u003cstrong\u003e69\u003c/strong\u003e, 610\u0026ndash;614 (2014).\u003c/li\u003e\n\u003cli\u003eLiu, H., Wang, Z., Li, Y., Li, W. \u0026amp; Chen, Y. Simvastatin prevents proliferation and bone metastases of lung adenocarcinoma in vitro and in vivo. \u003cem\u003eneo\u003c/em\u003e \u003cstrong\u003e60\u003c/strong\u003e, 240\u0026ndash;246 (2013).\u003c/li\u003e\n\u003cli\u003eYao, J., Zhang, Y., Ramishetti, S., Wang, Y. \u0026amp; Huang, L. Turning an antiviral into an anticancer drug: Nanoparticle delivery of acyclovir monophosphate. \u003cem\u003eJournal of Controlled Release\u003c/em\u003e \u003cstrong\u003e170\u003c/strong\u003e, 414\u0026ndash;420 (2013).\u003c/li\u003e\n\u003cli\u003ePezaro, C. PARP inhibitor combinations in prostate cancer. \u003cem\u003eTher Adv Med Oncol\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 1758835919897537 (2020).\u003c/li\u003e\n\u003cli\u003eVerveris, K. \u0026amp; Karagiannis, T. C. Potential non-oncological applications of histone deacetylase inhibitors. \u003cem\u003eAm J Transl Res\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 454\u0026ndash;467 (2011).\u003c/li\u003e\n\u003cli\u003eWu, C. W. \u003cem\u003eet al.\u003c/em\u003e Regulation of arginase production by glucocorticoid in three human gastric cancer cell lines. \u003cem\u003eLife Sci\u003c/em\u003e \u003cstrong\u003e51\u003c/strong\u003e, 1355\u0026ndash;1361 (1992).\u003c/li\u003e\n\u003cli\u003eBusada, J. T. \u003cem\u003eet al.\u003c/em\u003e Endogenous glucocorticoids prevent gastric metaplasia by suppressing spontaneous inflammation. \u003cem\u003eJ Clin Invest\u003c/em\u003e \u003cstrong\u003e129\u003c/strong\u003e, 1345\u0026ndash;1358 (2019).\u003c/li\u003e\n\u003cli\u003eTanaka, A. \u003cem\u003eet al.\u003c/em\u003e Role of copper ion in the pathogenesis of type 2 diabetes. \u003cem\u003eEndocr J\u003c/em\u003e \u003cstrong\u003e56\u003c/strong\u003e, 699\u0026ndash;706 (2009).\u003c/li\u003e\n\u003cli\u003eNagai, R., Murray, D. B., Metz, T. O. \u0026amp; Baynes, J. W. Chelation: a fundamental mechanism of action of AGE inhibitors, AGE breakers, and other inhibitors of diabetes complications. \u003cem\u003eDiabetes\u003c/em\u003e \u003cstrong\u003e61\u003c/strong\u003e, 549\u0026ndash;559 (2012).\u003c/li\u003e\n\u003cli\u003eChenoufi, N. \u003cem\u003eet al.\u003c/em\u003e Antiproliferative effect of deferiprone on the Hep G2 cell line. \u003cem\u003eBiochem Pharmacol\u003c/em\u003e \u003cstrong\u003e56\u003c/strong\u003e, 431\u0026ndash;437 (1998).\u003c/li\u003e\n\u003cli\u003eKontoghiorghes, G. J., Efstathiou, A., Ioannou-Loucaides, S. \u0026amp; Kolnagou, A. Chelators controlling metal metabolism and toxicity pathways: applications in cancer prevention, diagnosis and treatment. \u003cem\u003eHemoglobin\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 217\u0026ndash;227 (2008).\u003c/li\u003e\n\u003cli\u003eHawash, M. Recent Advances of Tubulin Inhibitors Targeting the Colchicine Binding Site for Cancer Therapy. \u003cem\u003eBiomolecules\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 1843 (2022).\u003c/li\u003e\n\u003cli\u003ePeerzada, M. N., Dar, M. S. \u0026amp; Verma, S. Development of tubulin polymerization inhibitors as anticancer agents. \u003cem\u003eExpert Opin Ther Pat\u003c/em\u003e \u003cstrong\u003e33\u003c/strong\u003e, 797\u0026ndash;820 (2023).\u003c/li\u003e\n\u003cli\u003eDimas, K. \u003cem\u003eet al.\u003c/em\u003e Steroidal cardiac Na+/K+ ATPase inhibitors exhibit strong anti-cancer potential in vitro and in prostate and lung cancer xenografts in vivo. \u003cem\u003eAnticancer Agents Med Chem\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 762\u0026ndash;770 (2014).\u003c/li\u003e\n\u003cli\u003eJoshi, S. K., Mikusa, J. P., Weaver, B. \u0026amp; Honore, P. Morphine and ABT-594 (a nicotinic acetylcholine agonist) exert centrally mediated antinociception in the rat cyclophosphamide cystitis model of visceral pain. \u003cem\u003eJ Pain\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 146\u0026ndash;156 (2008).\u003c/li\u003e\n\u003cli\u003ePalea, S., Artibani, W., Ostardo, E., Trist, D. G. \u0026amp; Pietra, C. Evidence for purinergic neurotransmission in human urinary bladder affected by interstitial cystitis. \u003cem\u003eJ Urol\u003c/em\u003e \u003cstrong\u003e150\u003c/strong\u003e, 2007\u0026ndash;2012 (1993).\u003c/li\u003e\n\u003cli\u003eGosis, B. S. \u003cem\u003eet al.\u003c/em\u003e Inhibition of nonalcoholic fatty liver disease in mice by selective inhibition of mTORC1. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e376\u003c/strong\u003e, eabf8271 (2022).\u003c/li\u003e\n\u003cli\u003eFeng, J. \u003cem\u003eet al.\u003c/em\u003e mTOR: A Potential New Target in Nonalcoholic Fatty Liver Disease. \u003cem\u003eInt J Mol Sci\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 9196 (2022).\u003c/li\u003e\n\u003cli\u003eDeng, T. \u003cem\u003eet al.\u003c/em\u003e Comparison of Genotype Imputation for SNP Array and Low-Coverage Whole-Genome Sequencing Data. \u003cem\u003eFront Genet\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 704118 (2021).\u003c/li\u003e\n\u003cli\u003eAn, J. \u003cem\u003eet al.\u003c/em\u003e Blood transcriptome differentiates clinical clusters for asthma. \u003cem\u003eWorld Allergy Organ J\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 100871 (2024).\u003c/li\u003e\n\u003cli\u003eLee, J. \u0026amp; Gim, J. Unlocking the Potential of Low Quality Total RNA-seq Data: A Stepwise Mapping Approach for Improved Quantitative Analyses. 2024.09.30.615750 Preprint at https://doi.org/10.1101/2024.09.30.615750 (2024).\u003c/li\u003e\n\u003cli\u003eMoon, S. \u003cem\u003eet al.\u003c/em\u003e The Korea Biobank Array: Design and Identification of Coding Variants Associated with Blood Biochemical Traits. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 1382 (2019).\u003c/li\u003e\n\u003cli\u003eSeo, S. \u003cem\u003eet al.\u003c/em\u003e SNP genotype calling and quality control for multi-batch-based studies. \u003cem\u003eGenes Genomics\u003c/em\u003e \u003cstrong\u003e41\u003c/strong\u003e, 927\u0026ndash;939 (2019).\u003c/li\u003e\n\u003cli\u003eSong, Y. E. \u003cem\u003eet al.\u003c/em\u003e ONETOOL for the analysis of family-based big data. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 2851\u0026ndash;2853 (2018).\u003c/li\u003e\n\u003cli\u003eHowie, B. N., Donnelly, P. \u0026amp; Marchini, J. A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. \u003cem\u003ePLoS Genet\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, e1000529 (2009).\u003c/li\u003e\n\u003cli\u003eYoo, S.-K. \u003cem\u003eet al.\u003c/em\u003e NARD: whole-genome reference panel of 1779 Northeast Asians improves imputation accuracy of rare and low-frequency variants. \u003cem\u003eGenome Med\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 64 (2019).\u003c/li\u003e\n\u003cli\u003eLoh, P.-R. \u003cem\u003eet al.\u003c/em\u003e Reference-based phasing using the Haplotype Reference Consortium panel. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 1443\u0026ndash;1448 (2016).\u003c/li\u003e\n\u003cli\u003eFuchsberger, C., Abecasis, G. R. \u0026amp; Hinds, D. A. minimac2: faster genotype imputation. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 782\u0026ndash;784 (2015).\u003c/li\u003e\n\u003cli\u003eBolger, A. M., Lohse, M. \u0026amp; Usadel, B. Trimmomatic: a flexible trimmer for Illumina sequence data. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 2114\u0026ndash;2120 (2014).\u003c/li\u003e\n\u003cli\u003eDobin, A. \u003cem\u003eet al.\u003c/em\u003e STAR: ultrafast universal RNA-seq aligner. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 15\u0026ndash;21 (2013).\u003c/li\u003e\n\u003cli\u003eDeLuca, D. S. \u003cem\u003eet al.\u003c/em\u003e RNA-SeQC: RNA-seq metrics for quality control and process optimization. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 1530\u0026ndash;1532 (2012).\u003c/li\u003e\n\u003cli\u003eHarrow, J. \u003cem\u003eet al.\u003c/em\u003e GENCODE: the reference human genome annotation for The ENCODE Project. \u003cem\u003eGenome Res\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 1760\u0026ndash;1774 (2012).\u003c/li\u003e\n\u003cli\u003eRobinson, M. D. \u0026amp; Oshlack, A. A scaling normalization method for differential expression analysis of RNA-seq data. \u003cem\u003eGenome Biol\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, R25 (2010).\u003c/li\u003e\n\u003cli\u003eStegle, O., Parts, L., Piipari, M., Winn, J. \u0026amp; Durbin, R. Using probabilistic estimation of expression residuals (PEER) to obtain increased power and interpretability of gene expression analyses. \u003cem\u003eNat Protoc\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 500\u0026ndash;507 (2012).\u003c/li\u003e\n\u003cli\u003eGTEx Consortium \u003cem\u003eet al.\u003c/em\u003e Genetic effects on gene expression across human tissues. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e550\u003c/strong\u003e, 204\u0026ndash;213 (2017).\u003c/li\u003e\n\u003cli\u003eFriedman, J., Hastie, T. \u0026amp; Tibshirani, R. Regularization Paths for Generalized Linear Models via Coordinate Descent. \u003cem\u003eJ Stat Softw\u003c/em\u003e \u003cstrong\u003e33\u003c/strong\u003e, 1\u0026ndash;22 (2010).\u003c/li\u003e\n\u003cli\u003eShabalin, A. A. Matrix eQTL: ultra fast eQTL analysis via large matrix operations. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 1353\u0026ndash;1358 (2012).\u003c/li\u003e\n\u003cli\u003eYang, J., Lee, S. H., Goddard, M. E. \u0026amp; Visscher, P. M. GCTA: a tool for genome-wide complex trait analysis. \u003cem\u003eAm J Hum Genet\u003c/em\u003e \u003cstrong\u003e88\u003c/strong\u003e, 76\u0026ndash;82 (2011).\u003c/li\u003e\n\u003cli\u003eKuznetsova, I., Lugmayr, A., Siira, S. J., Rackham, O. \u0026amp; Filipovska, A. CirGO: an alternative circular way of visualising gene ontology terms. \u003cem\u003eBMC Bioinformatics\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 84 (2019).\u003c/li\u003e\n\u003cli\u003eKnox, C. \u003cem\u003eet al.\u003c/em\u003e DrugBank 6.0: the DrugBank Knowledgebase for 2024. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, D1265\u0026ndash;D1275 (2024).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6417034/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6417034/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eGenome-wide association studies (GWAS) have linked genetic loci to complex traits, but their interpretation is limited by linkage disequilibrium (LD) and difficulties in identifying causal genes. Transcriptome-wide association studies (TWAS) address these challenges by leveraging predicted gene expression to map genomic risk regions. However, TWAS reference panels are predominantly based on European ancestry, restricting their utility for other populations. Here, we developed a Korean-specific whole blood reference panel for transcriptome imputation and performed TWAS, examining 81 traits in 79,294 Korean individuals. This Korean panel showed improved predictive performance over the GTEx reference panel, identifying 348 significant gene-trait associations (Bonferroni-corrected \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), including 232 novel associations, compared to 180 associations with GTEx. Furthermore, pathway analysis revealed unique biological processes in the Korean model, offering insights into diseases such as thyroid disease and type 2 diabetes. Drug repurposing identified promising candidates, including clopidogrel for angina and vorinostat for fatty liver. These findings highlight the value of population-specific approaches in genetic research.\u003c/p\u003e","manuscriptTitle":"Transcriptome-Wide Association Studies of 81 traits in 79,294 Korean individuals","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-17 09:28:22","doi":"10.21203/rs.3.rs-6417034/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"25644793-cca4-4640-9bc0-dc28a4e04328","owner":[],"postedDate":"April 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":47259619,"name":"Biological sciences/Genetics/Genetic association study"},{"id":47259620,"name":"Biological sciences/Genetics/Gene expression"},{"id":47259621,"name":"Biological sciences/Genetics/Functional genomics/Gene expression profiling"},{"id":47259622,"name":"Biological sciences/Genetics/Genomics/Transcriptomics"}],"tags":[],"updatedAt":"2025-06-17T18:20:18+00:00","versionOfRecord":[],"versionCreatedAt":"2025-04-17 09:28:22","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6417034","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6417034","identity":"rs-6417034","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}