Integrative Multi-Omics Framework for Causal Gene Discovery in Long COVID

Integrative Multi-Omics Framework for Causal Gene Discovery in Long COVID | medRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (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];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-P4HH5NV'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search Integrative Multi-Omics Framework for Causal Gene Discovery in Long COVID View ORCID Profile Sindy Pinero , View ORCID Profile Xiaomei Li , View ORCID Profile Lin Liu , View ORCID Profile Jiuyong Li , View ORCID Profile Sang Hong Lee , View ORCID Profile Marnie Winter , View ORCID Profile Thin Nguyen , View ORCID Profile Junpeng Zhang , View ORCID Profile Thuc Duy Le doi: https://doi.org/10.1101/2025.02.09.25321751 Sindy Pinero 1 *UniSA STEM, University of South Australia , Mawson Lakes, Adelaide, 5095, SA, Australia . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Sindy Pinero Xiaomei Li 2 Agriculture and Food, Commonwealth Scientific and Industrial Research Organisation , 26 Pembroke Rd, Marsfield, 2122, NSW, Australia . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Xiaomei Li Lin Liu 1 *UniSA STEM, University of South Australia , Mawson Lakes, Adelaide, 5095, SA, Australia . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Lin Liu Jiuyong Li 1 *UniSA STEM, University of South Australia , Mawson Lakes, Adelaide, 5095, SA, Australia . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Jiuyong Li Sang Hong Lee 3 Australian Centre for Precision Health, University of South Australia , Adelaide, SA, 5000, Australia . 4 UniSA Allied Health and Human Performance, University of South Australia , Adelaide, SA, 5000, Australia . 5 South Australian Health and Medical Research Institute (SAHMRI), University of South Australia , Adelaide, SA, 5000, Australia . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Sang Hong Lee Marnie Winter 6 Future Industries Institute, University of South Australia , Mawson Lakes, Adelaide, 5095, SA, Australia . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Marnie Winter Thin Nguyen 7 Applied Artificial Intelligence Institute, Deakin University , 75 Pigdons Rd, Victoria, 3216, Melbourne, Australia . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Thin Nguyen Junpeng Zhang 8 School of Engineering, Dali University , 2 Hongsheng Rd, Dali, 671003, Yunnan, China . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Junpeng Zhang Thuc Duy Le 1 *UniSA STEM, University of South Australia , Mawson Lakes, Adelaide, 5095, SA, Australia . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Thuc Duy Le For correspondence: Thuc.Le{at}unisa.edu.au Abstract Full Text Info/History Metrics Data/Code Preview PDF Abstract Background Long COVID, or Post-Acute Sequelae of COVID-19 (PASC), involves persistent, multisystemic symptoms in about 10–20% of COVID-19 patients. Although age, sex, ethnicity, and comorbidities are recognized as risk factors, identifying genetic contributors is essential for developing targeted therapies. Methods We developed a multi-omics framework using Transcriptome-Wide Mendelian Randomization (TWMR) and Control Theory (CT). This approach integrates Expression Quantitative Trait Loci (eQTL), Genome-Wide Association Studies (GWAS), RNA sequencing (RNA-seq), and Protein-Protein Interaction (PPI) networks to detect causal genes and regulatory nodes that drive critical expression changes in Long COVID. Results We identified 32 causal genes (19 previously reported and 13 novel), which act as regulatory drivers influencing disease risk, progression, and stability. Enrichment analyses highlighted pathways linked to the SARS-CoV-2 response, viral carcinogenesis, cell cycle regulation, and immune function. Analysis of other pathophysiological conditions revealed shared genetic factors across syndromic, metabolic, autoimmune, and connective tissue disorders. Using these genes, we identified three distinct symptom-based subtypes of Long COVID, offering insights for more precise diagnosis and potential therapeutic interventions. Additionally, we provided an open-source Shiny application to enable further data exploration. Conclusion Integrating TWMR and CT revealed genetic mechanisms and therapeutic targets for Long COVID, with novel genes informing pathogenesis and precision medicine strategies. 1 Introduction Long COVID, also known as Post-Acute Sequelae of COVID-19 (PASC), is a complex condition marked by the persistence or development of symptoms as a result of SARS-CoV-2 infection. Long COVID is defined differently by various organizations. For instance, the World Health Organization (WHO) and the Centers for Disease Control and Prevention (CDC) describe it as symptoms persisting three months after infection and lasting for at least two months [ 1 , 2 ], whereas the National Institute for Health and Care Excellence (NICE) considers it to begin as early as one-month post-infection [ 3 – 5 ]. Regardless of the definition timing, key risk factors include age, sex, ethnicity, socioeconomic status, vaccination status, smoking, and underlying health conditions [ 6 ]. Moreover, studies have linked various biomarkers to Long COVID, particularly those related to inflammation, immune dysfunction, and coagulation abnormalities [ 7 ]. Despite significant progress in identifying risk factors and clinical markers [ 6 , 7 ], understanding the role of gene expression as a causal factor in Long COVID remains a major challenge. This gap in knowledge presents a significant barrier to developing and implementing interventions and targeted therapies [ 8 ], highlighting the need for novel approaches focusing on gene expression patterns that may contribute to Long COVID. Identifying these Long COVID-causing genes is essential for advancing targeted treatment strategies. It could also improve diagnostic accuracy and promote better monitoring and prediction of patient outcomes [ 9 ]. Computational methods for identifying disease-causing genes typically follow two primary strategies, each offering distinct advantages that complement the other. The first strategy aims to identify genes associated with disease risk and prevention, often using approaches such as Transcriptome-Wide Mendelian Randomization (TWMR) [ 10 ]. The TWMR approach incorporates transcriptomic data into MR studies by using genetic variants that affect gene expression, such as Expression Quantitative Trait Loci (eQTLs), to establish causal relationships between gene activities and diseases. The TWMR methodology can identify whether altered gene expression directly influences an outcome—in this instance, Long COVID—and reveal potential therapeutic targets. The resulting analysis reveals which genetic factors influence disease susceptibility or protection through genetic associations and causal inference, allowing researchers to identify specific genetic variants with direct causal effects on diseases. However, TWMR analysis often requires strong genetic instruments (e.g., Single Nucleotide Polymorphisms (SNPs) that robustly modulate gene expression), and determining causal relationships becomes more complex when facing confounding variables or pleiotropy. The second strategy identifies genes or proteins that are crucial in biological networks. Techniques such as Bayesian Networks [ 11 ], Node Importance [ 12 ], and the Control Theory (CT) [ 13 ] are used to understand how different genes and proteins interact within biological pathways, considering the interconnected nature of biological systems. CT is particularly useful for identifying critical nodes or key genes and proteins significantly influencing the entire network. By finding these critical nodes (network driver genes), researchers can determine which components are most effective to target to stabilize or control disease-related disruptions. For example, CT methods have been used in cancer research to identify key regulatory genes like TP53 whose modulation can restore network stability, thereby providing focused therapeutic opportunities [ 14 ]. In this study, we propose a novel framework to explore and discover potential genes involved in Long COVID by integrating two complementary strategies: MR [ 10 ] and CT [ 13 ], along with multi-omics data, including eQTLs, Genome-Wide Association Studies (GWAS), RNA sequencing (RNA-seq), and human Protein-Protein Interaction (PPI) network. Our framework investigates candidate causal genes that may contribute to Long COVID risk and examines their potential regulatory roles within a network. Specifically, we discover genes whose expression patterns suggest either increased susceptibility to Long COVID or a crucial role in maintaining biological network stability. By integrating these methodologies and utilizing multi-omics data, our analysis provides comprehensive insights into the potential genetic mechanisms underlying Long COVID and highlights candidate therapeutic targets for further investigation. Our study has identified 32 causal genes for Long COVID, of which 19 have been confirmed by existing literature, supporting the effectiveness of our findings. The remaining genes represent promising candidates for follow-up experiments. Among these candidates, we discovered driver genes that act as risk or preventive factors and network driver genes that regulate and stabilize disease network structure. Enrichment analyses revealed important biological pathways in Long COVID, including SARS-CoV-2 infection, viral carcinogenesis, cell cycle regulation, and immune response mechanisms. Using the identified causal genes, we clustered Long COVID patients into three distinct subtypes with different symptom profiles, establishing a foundation for personalized diagnostic and therapeutic approaches. This work represents a significant step toward customized management and treatment strategies for Long COVID, ultimately improving patient outcomes. To facilitate the application of our framework, we developed a web application (Shiny app) that allows users to generate gene lists by adjusting parameters related to direct (MR) and network-based (CT) causal approaches. This tool provides researchers and clinicians with an accessible platform to explore parameter variations and analyze their data, enhancing the reproducibility of our findings. 2 Results 2.1 Overview of the Causal Gene Discovery Framework The causal gene discovery framework integrates diverse data sources, including eQTL, GWAS, RNA-seq, and PPI networks, to identify genes with causal roles in Long COVID ( Fig. 1 ). It begins by processing multi-omics input data ( Fig. 1A ) and then applies an integrative scoring method ( Fig. 1B ) that combines TWMR with network analysis using CT. This approach balances the contributions of risk/preventive factors and network-critical genes through a parameter ( α ) that can be adjusted depending on the goals. The output ( Fig. 1C ) ranks causal genes based on the weighted scores, offering insights into their roles within the Long COVID network. Finally, downstream analyses ( Fig. 1D ), including Enrichment Analysis (EA), literature validation, and subtype identification, help discover disease mechanisms and prioritize therapeutic targets. This comprehensive computational approach integrates genetic and network-based perspectives, providing deeper insights into the nature of Long COVID. Download figure Open in new tab Fig. 1: A causal gene discovery framework for Long COVID using multi-omics data. (A) The input data includes expression Quantitative Trait Loci (eQTL), Long COVID Genome-Wide Association Studies (GWAS), RNA sequencing (RNA-seq), and the human Protein-Protein Interaction (PPI) network. (B) A fusion approach to evaluating gene expression by integrating Transcriptome-Wide Mendelian Randomization (TWMR) and Control Theory (CT) scores. (C) Significant genes are ranked based on the weighted scores. (D) Downstream analyses include Enrichment Analysis (EA), literature review, and Long COVID subtype identification. SNPs: Single Nucleotide Polymorphisms, IVs: Instrumental Variables. E: Exposure. O: Outcome. U: Confounders. By treating genetic variants as instrumental variables (IVs), two-sample MR methods detect genetically regulated risk exposures for complex diseases using only summary statistics. When considering gene expression as an exposure in TWMR analyses, we aim to identify gene expressions that have causal relationships with the disease of interest. In our case, we focus on identifying the genes that act as risk or protective factors for Long COVID. Given the limited number of eQTLs available as IVs for a gene, which makes detecting invalid IVs challenging, we adopt the multi-tissue approach, Mt-Robin [ 10 ]. This method uses eQTL data in a mixed model to identify IV-specific random effects due to pleiotropy arising from estimation errors in eQTL summary statistics, enabling accurate inference of the dependence (fixed effects) between eQTL and GWAS effects, even in the presence of invalid IVs. While MR approaches identify genes that directly affect Long COVID, network biology approaches, such as CT, have shown that genes driving the disease are not limited to those directly linked to disease phenotypes [ 13 ]. In this work, we employ a CT approach to extract a list of genes that serve as network drivers for Long COVID—i.e., genes whose removal or intervention would disrupt the biological networks associated with the disease, thereby affecting disease outcomes. These network driver genes may or may not have direct causal relationships with the disease. To create a comprehensive list of causal genes for Long COVID and understand their roles in regulating the disease, we use a fusion approach that integrates the two methods described above (see Methods for details). Specifically, we calculate the scores of each gene using the following formula: where: S Causal represents the final score of each gene. S Risk ( D GWAS , D eQTL ) is the score derived from the TWMR approach ( Mt-Robin ) to identify risk and protective causal genes using GWAS and eQTL datasets. S Network ( D RNA-seq , D PPI ) is the score from the CT approach to identify network driver genes based on RNA-seq data and the human PPI network. The parameter α controls the contribution of each risk/protective causal gene, while 1 − α adjusts the influence of each network-critical gene. The formula (1) integrates an approach that ranks genes by combining their causal effects and significance within the Long COVID network, providing a comprehensive prioritization based on both causal and network properties. Thus, this causal multi-omics approach provides insights into Long COVID’s genetic mechanisms while highlighting possible intervention targets. 2.2 Dynamic Visualization of Long COVID Causal Genes: A Shiny Application In our model, the parameter α serves as an adjustable coefficient that enables researchers to explore different scenarios for prioritizing protein-coding genes based on their roles in influencing disease risk or prevention and controlling the Long COVID network. When α approaches 1, the model prioritizes genes linked to disease risk and preventive scenarios. These genes are directly associated with Long COVID pathogenesis, highlighting potential therapeutic targets for intervention. Conversely, as α approaches 0, the model emphasizes driver genes critical to the disease network’s structure. These genes regulate key interactions within the network, positioning them as potential therapeutic targets to restore lost stability or modulate pathological states. The model integrates both perspectives in the intermediate ranges of α , balancing the importance of network controllability with disease risk. In these cases, genes that significantly influence the network and are closely linked to disease risk become key players, making them important targets for further investigation. Researchers can dynamically explore these shifts in gene rankings by adjusting α in our interactive tool available at Dynamic Causal Genes Visualization in a Long COVID Network 1 . This instrument allows a detailed examination of how genes transition from disease risk or preventive factors ( α → 1) to network drivers ( α → 0). 2.3 Causal Genes of Long COVID By varying α values in our model, we identified a comprehensive set of causal genes for Long COVID with distinct roles. Fig. 2 shows the sets of these causal genes corresponding to specific values of α . As α approaches 1, the model outputs genes classified as risk (red) or preventive (green), inferred from the color coding of their effect sizes, with red representing positive effect sizes (risk) and green representing negative effect sizes (preventive). Decreasing α towards zero shifts the focus to network driver genes that control the Long COVID PPI network (yellow). Download figure Open in new tab Fig. 2: Top causal genes ranked by their final score S Causal . These genes, obtained from our framework, are sorted horizontally based on their absolute effect size in ascending order and classified vertically across different α values. The parameter α balances the direct effect of genes on the disease ( S Risk ) and their network controllability roles ( S Network ). At α = 1, the model outputs disease risk (red) and prevention (green) genes, while decreasing α towards 0, the focus shifts to network driver genes that control the biological network (yellow). Genes like membrane occupation and recognition nexus repeat containing 4 ( MORN4 ), cell division cycle associated 26 ( CDC26 ), and eukaryotic translation initiation factor 5A ( EIF5A ) consistently rank highly across different α values (1.00 to 0.50), suggesting a strong causal relationship between their expression levels and disease risk or preventive mechanisms. Using SNPs as IVs in our analysis, we estimated the causal effects of the expression of these genes on Long COVID risk. The consistently high ranking of MORN4 , CDC26 , and EIF5A indicates that their expression levels may significantly contribute to disease susceptibility, making them potential key targets for intervention strategies focused on reducing disease risk (see Fig. 2 ). The complete list of SNPs used as IVs for each gene’s expression can be found in Supplementary Data 1 . As α decreases, the model shifts focus from the MR perspective to the CT perspective, prioritizing the balance between risk-related genetic contributions and network control dynamics. This transition highlights the framework’s flexibility in integrating these perspectives. Notably, genes such as tumor protein p53 ( TP53 ), cyclic adenosine monophosphate response element-binding protein-binding-protein ( CREBBP ), early region 1A binding protein p300 ( EP300 ), tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein gamma ( YWHAG ), SMAD family member 3 ( SMAD3 ), and the growth factor receptor-bound protein 2 ( GRB2 ) become increasingly crucial in the network, emphasizing their roles in maintaining network control (see Fig. 2 , with these genes highlighted in yellow). When considering the union of the top genes for each α value in our analysis, we identified 32 unique core causal genes for Long COVID. This comprehensive set of genes represents the most influential factors across the spectrum of our α parameter, which balances disease-related impact and network controllability. Of these 32 genes, 19 have been previously identified in COVID-19 and/or Long COVID studies, reinforcing their importance in the disease process. These include well-known genes such as the androgen receptor ( AR ), butyrophilin subfamily 3 member A1 ( BTN3A1 ), cyclin-dependent kinase inhibitor 1A ( CDKN1A ), CREBBP , EIF5A , EP300 , estrogen receptor 1 ( ESR1 ), atos homolog A ( ATOSA ), FYN proto-oncogene ( FYN ), GRB2 , histone deacetylase 1 ( HDAC1 ), mitogen-activated protein kinase 1 ( MAPK1 ), NADH:ubiquinone oxidoreductase subunit A6 ( NDUFA6 ), retinoblastoma transcriptional corepressor 1 ( RB1 ), SMAD family member 2 ( SMAD2 ), SMAD3 , sarcoma proto-oncogene ( SRC ), TP53 , and YWHAG . These genes have been associated with various SARS-CoV-2 infection and Long COVID aspects, including roles as hub genes, drug targets, and factors influencing disease severity ( Table 1 ). The high number of confirmed Long COVID genes suggests that our framework effectively identifies causal genes. View this table: View inline View popup Download powerpoint Table 1: Core causal genes for Long COVID confirmed by the literature. These 19 genes were validated by existing COVID-19 (COV) and/or Long COVID (LCV) studies, reinforcing our findings. Sev.: Severity, Reg.: Regulation, Polymo.: Polymorphisms. For more supporting literature, refer to Supplementary Data 2 . The remaining 13 genes in our causal gene set are novel discoveries in the context of COVID-19 and Long COVID research: adenosine deaminase tRNA-specific 1 ( ADAT1 ), B-cell lymphoma 2 interacting protein 1 ( BNIP1 ), bole-like 2 ( BOLA2 ), chromosome 19 open reading frame 18 ( C19orf18 ), inositol 1,4,5-trisphosphate receptor interacting domain containing 1 ( ITPRID1 ), CDC26 , cytidine deaminase ( CDA ), ceramide synthase 4 ( CERS4 ), casein kinase 2 alpha 1 ( CSNK2A1 ), GDP-mannose pyrophosphorylase B synthase ( GMPPB ), MORN repeat containing 3 ( MORN3 ), MORN4 and the von Willebrand factor D and EGF domains gene ( VWDE ). These genes have not been previously linked to COVID-19 or Long COVID, demonstrating the potential of our framework to reveal novel targets for intervention and further study. Enrichment analysis of the 32 causal genes identified 458 significant pathways in GO (Gene Ontology) [ 41 ], 99 in KEGG (Kyoto Encyclopedia of Genes and Genomes) [ 42 ], and 246 in Reactome [ 43 ]. The top 20 pathways from each database, ranked by adjusted p-value, are shown in Fig. 3 , with the complete list available in Supplementary Data 3 . Download figure Open in new tab Fig. 3: Enrichment analysis (EA) results for the identified Long COVID Causal Genes. (A) Gene Ontology (GO) EA, showing the top 20 enriched terms across Biological Process (BP) and Molecular Function (MF) categories. (B) KEGG pathway EA, displaying the top 20 enriched pathways. (C) Reactome pathway EA, illustrating the top 20 enriched pathways. For all plots, genes are ranked by the lowest adjusted p-value. The y-axis represents the enriched terms or pathways, the size of each dot reflects the number of associated genes, and the color gradient indicates the adjusted p-value, with red denoting greater significance. Key findings include the transforming growth factor (TGF)-Beta signaling pathway, highlighted in GO and KEGG analyses, which plays a crucial role in immune regulation and tissue repair. Its disruption may contribute to persistent inflammation and fibrosis, leading to lung and organ damage as observed in Long COVID patients [ 18 ]. Similarly, KEGG pathways like cell cycle and viral carcinogenesis suggest long-term cellular effects of SARS-CoV-2 infection, such as abnormal proliferation and senescence, potentially explaining prolonged recovery and tissue dysfunction [ 33 ]. GO analysis highlights the importance of immune signaling pathways in ongoing inflammation and autoimmune-like symptoms [ 44 ]. The Reactome analysis emphasizes Forkhead box O (FOXO) transcription and Phosphoinositide 3-kinase (PI3K)/Protein Kinase B (AKT) signaling, which are involved in metabolism, stress responses, cell survival, and growth factor signaling pathways that may impair tissue repair and regeneration [ 18 , 21 ]. These findings reveal potential mechanisms underlying Long COVID and suggest therapeutic targets, such as TGF-Beta signaling and FOXO transcription, to mitigate long-term effects. 2.3.1 Shared Genetic Basis of Long COVID and Related Conditions We examined the involvement of the 32 identified Long COVID causal genes in other pathophysiological conditions ( Table 2 , complete dataset in Supplementary Data 4 ). This analysis revealed several distinct patterns of disease overlap, curated from multiple disease databases, including The Human Disease Database (MalaCards), Disease-Gene Associations (DISEASES), The Gene-Disease Network (DisGeNET), Medical Genetics Database (MedGen), and the Gene Curation Coalition (GenCC) (see the Methods section for more information about these databases). Many of these genes are implicated in a spectrum of syndromic, metabolic, autoimmune, connective tissue, and neurodevelopmental disorders that share clinical or biological features with Long COVID manifestations [ 7 , 45 , 46 ]. View this table: View inline View popup Download powerpoint Table 2: Causal genes in Long COVID and their overlap with other pathophysiological conditions. Analysis of the identified causal genes for Long COVID shows their involvement in other pathophysiological conditions and potential mechanistic overlap with Long COVID manifestations. A subset of these genes ( CDKN1A , CREBBP , CSNK2A1 , and TP53 ) are involved in tumor-predisposition syndromes and complex developmental disorders with autoimmune and inflammatory components. These conditions’ aberrant cytokine signaling and dysregulated immune checkpoints suggest potential mechanisms for prolonged inflammatory responses observed in Long COVID [ 7 ]. Similarly, genes such as C19orf18 , CDC26 , MORN3 , NDUFA6 , VWDE , and YWHAG are linked to systemic conditions affecting multiple organ systems. Their association with mitochondrial dysfunction and vascular pathologies parallels the fatigue, dysautonomia, and endothelial dysfunction commonly reported in Long COVID [ 45 ]. ATOSA and GMPPB are linked to chronic inflammation, mirroring mechanisms of immune activation and tissue damage implicated in Long COVID. Additionally, CERS4 , ESR1 , FYN , and MAPK1 highlight the interplay between immune dysfunction and metabolic disruption, shedding light on the metabolic dysregulation seen in some patients [ 46 ]. Our database integration analysis reveals meaningful biological connections between Long COVID and other disorders, particularly immune-mediated conditions and metabolic diseases. The identified genetic overlaps suggest that variants in these genes might influence individual susceptibility to persistent post-viral symptoms, similar to their role in other chronic conditions. These shared molecular features help explain the diverse manifestations observed across Long COVID patients [ 47 ]. 2.3.2 Risk/Preventive Genes Driving Long COVID Susceptibility Among the 32 causal genes obtained from our framework, we identified 16 significant protein-coding genes directly associated with the risk and prevention of Long COVID. These genes are implicated in critical biological processes such as cell cycle regulation ( CDC26 ), apoptosis ( BNIP1 ), and immune response ( BTN3A1 ) ( Table 3 ). View this table: View inline View popup Download powerpoint Table 3: Risk and preventive causal genes for Long COVID ordered by the S Causal score. Genes are classified as Risk or Preventive factors for Long COVID based on their effect size sign (positive or negative, respectively) when α = 1. The forest plot ( Fig. 4 ) reveals a wide range of effect sizes for 16 protein-coding genes, from -30.04 for CDC26 to 34.22 for MORN4 , indicating varying degrees of influence on Long COVID susceptibility. Applying our framework, we identified statistically significant causal relationships for these genes (p-value and False Discovery Rate (FDR) < 0.05), with confidence intervals that do not cross zero, providing strong evidence for their potential roles. Notably, genes such as MORN4 , CDC26 , EIF5A , and VWDE exhibit the strongest causal associations, with the largest absolute effect sizes. Download figure Open in new tab Fig. 4: Effect size of the risk and preventive causal genes for Long COVID. Forest plot shows the significant genes identified at α = 1.0, with all causal relationships meeting statistical significance (p-value and FDR < 0.05). Higher expression is associated with increased (positive effect size) or decreased (negative effect size) risk. SNPs: number of associated SNPs; Tissues: number of tissues where the SNPs influence the gene expression. Points show fixed effect size (standardized beta coefficient) with 95% CI error bars. Red: lung and other tissues; Blue: non-lung tissues. GWAS: Genome-Wide Association Study. SNP: Single Nucleotide Polymorphism. FDR: False Discovery Rate. CI: Confidence Interval. In our framework, we used varying numbers of SNPs as IVs for each gene’s expression—ranging from 2 SNPs for genes like in MORN4 , CDC26 , EIF5A , GMPPB , and NDUFA6 , to 18 SNPs for MORN3 . These IVs strengthen the validity of our causal estimates between gene expression and Long COVID risk. The number of the tissues in which the gene expression was evaluated also varied by gene, enhancing the robustness of our findings across different biological contexts. For instance, MORN4 showed expression changes in two tissues/cells (left ventricle and thyroid) and in cultured fibroblasts. In contrast, CDA exhibited a broad impact, affecting gene expression across 26 distinct tissues. These span multiple systems: adipose, brain, cardiovascular, endocrine, connective, immune, digestive, reproductive, excretory tissues, and blood. This extensive tissue distribution highlights the far-reaching effects of SNPs on gene expression throughout the body. Moreover, the expression patterns of all the 16 risk/preventive protein-coding genes identified through our framework suggest a systemic involvement in Long COVID. Ten genes showed expression in both lung and other tissues, while six genes were expressed exclusively in non-lung tissues. This distribution of expression patterns in other non-lung tissues supports the presence of non-respiratory symptoms observed in Long COVID patients, which suggests the involvement of molecular mechanisms beyond the pulmonary system [ 48 ]. The directional effects vary among the genes, with some showing positive effect sizes (e.g., ATOSA , BNIP1 , C19orf18 , EIF5A , GMPPB , MORN3 , MORN4 , NDUFA6 , and VWDE ) and others negative effect sizes (e.g., ADAT1 , BOLA2 , BTN3A1 , CDA , CDC26 , CERS4 , and ITPRID1 ). Genes with positive effect sizes suggest increased expression in relevant tissues is associated with higher Long COVID susceptibility. In contrast, those with negative effect sizes indicate that increased expression may reduce the risk or be protective against Long COVID. Among these 16 protein-coding genes, the roles of BTN3A1 , EIF5A , and NDUFA6 were previously identified in the pathogenesis of COVID-19, suggesting a potential link between their expression and the development of Long COVID. [ 17 , 22 , 31 ] ( Table 4 ). View this table: View inline View popup Download powerpoint Table 4: Summary of three causal genes with established links to COVID-19 and hypothesized effects in Long COVID. COVID-19: Coronavirus Disease 2019. PRF: Programmed Ribosomal Frameshifting. Additional related literature and references can be found in the Supplementary Data 2 . BTN3A1 , an immune system protein involved in T-cell activation and regulation, is part of a 5-gene signature predicting ventilator-free days in COVID-19 patients [ 17 ]. Our analysis revealed a negative effect size value for BTN3A1 , suggesting that higher expression is causally associated with better clinical outcomes and potentially reduced Long COVID risk. This protective effect may be attributed to its role in promoting a more controlled immune response, thereby reducing long-term complications [ 49 ]. Conversely, EIF5A , a translation factor that promotes programmed ribosomal frameshifting (PRF), translation termination, and ribosome recycling in SARS-CoV-2 infection, showed a positive effect size value. This function indicates that EIF5A could contribute to persistent symptoms in Long COVID through ongoing disruption in translation regulation and protein synthesis, resulting in continued immune activation and cellular stress [ 22 ]. NDUFA6 , a key component of the mitochondrial respiratory chain, has been identified among the top genes associated with SARS-CoV-2 infection [ 31 ], showing significant mRNA expression differences in affected patients. Our findings indicate that disruptions in NDUFA6 may adversely affect cardiovascular health and elevate Long COVID risk. These effects are likely attributable to the gene’s critical role in cellular energy production and mitochondrial function. Impaired activity of NDUFA6 can lead to reduced ATP synthesis, increased oxidative stress, and the development of cardiovascular symptoms frequently observed in Long COVID patients [ 50 ]. These findings suggest that BTN3A1 , EIF5A , and NDUFA6 play significant roles in COVID-19 and may have implications in Long COVID, with BTN3A1 potentially mitigating Long COVID risk through improved immune regulation, while EIF5A and NDUFA6 contributing to persistent symptoms due to their roles in viral response and energy production, respectively. In addition to the three previously mentioned genes, our framework identified 13 novel risk/preventive causal genes for Long COVID. Among these genes, CDA , ADAT1 , CERS4 , CDC26 , and BOLA2 were mainly enriched in our analyses with significant roles in crucial pathways, including nucleotide metabolism, RNA editing, lipid metabolism, cell cycle regulation, and iron-sulfur cluster assembly. CDA and ADAT1 are both involved in nucleotide metabolism and RNA editing processes. CDA is crucial for pyrimidine salvage and nucleotide pool balance, potentially impacting RNA integrity and immune system function [ 53 ]. Similarly, ADAT1 is implicated in pre-mRNA editing, converting adenosine to inosine in eukaryotic tRNA, potentially influencing inflammatory responses [ 54 ]. Their roles as risk factors may be hypothesized based on their involvement in these critical cellular processes, which could contribute to the persistent symptoms observed in Long COVID patients [ 4 ]. CERS4 and BOLA2 are involved in cellular metabolism and homeostasis. CERS4 facilitates sphingosine N-acyltransferase activity and is implicated in ceramide synthesis, influencing lipid metabolism and cellular signaling pathways [ 55 ]. BOLA2 works in iron maturation and is part of the iron-sulfur cluster assembly complex, playing a role in cell redox homeostasis [ 56 ]. The risk association of these genes might be related to their impact on various cellular processes, including signaling pathways and cellular respiration. Its role may be associated with the diverse symptoms observed in Long COVID cases [ 57 ]. CDC26 is part of the anaphase-promoting complex (APC) involved in cell cycle regulation [ 58 ]. Its role as a risk factor may be attributed to its function as a ubiquitin-protein ligase, managing the proteolysis of cell cycle proteins, which could impair cellular repair and regeneration processes, possibly explaining the prolonged cellular damage found in individuals with Long COVID [ 59 ]. Comprehensive results of pathway enrichment analyses using GO, KEGG, and Reactome databases, including significantly enriched biological processes, molecular functions, cellular components, and pathways, are detailed in Supplementary Data 5 . 2.3.3 Network Driver Genes Controlling Long COVID Network There were 16 genes in the causal gene list found as network drivers of Long COVID. In the context of CT, network driver genes are critical nodes within a biological network whose manipulation can control the overall state and dynamics of the system. They serve as key regulators that can influence the activity of numerous downstream genes and pathways. Targeting these driver genes makes it possible to control the network toward a desired state, such as restoring normal function or mitigating disease effects. By identifying these genes, our framework highlights key intervention points within the network, offering potential targets to influence and modify its behavior and outcomes [ 13 ]. These identified core network driver genes have at least 150 connections with other nodes, emphasizing their significant influence. Disruption of these genes under normal conditions could contribute to the pathogenesis of Long COVID, making them potential therapeutic targets for restoring normal function in affected patients ( Table 5 ). View this table: View inline View popup Download powerpoint Table 5: Network driver genes for Long COVID ordered by the S Causal score. The K column represents the total degree (total number of interactions), K in describes the in-degree (incoming interactions), and K out denotes the out-degree (outgoing interactions). Of the 16 network driver genes identified, 14 were associated with enriched pathways crucial in COVID-19, Long COVID, or both. These pathways involved essential cellular functions such as cell proliferation, differentiation, cycle progression, DNA repair, inflammation, and immune responses. Disruptions in these processes can lead to the persistent symptoms of Long COVID, chronic inflammation, neurodegeneration, and immune dysfunction ( Table 6 ). View this table: View inline View popup Download powerpoint Table 6: Long COVID roles of the identified network driver genes. Key protein functions and enriched pathways obtained from GO, KEGG, or Reactome are shown, along with their roles in COVID-19 and Long COVID pathogenesis, illustrating their contribution to disease mechanisms. All pathway enrichments meet statistical significance thresholds (p-value and FDR < 0.05). FDR: False Discovery Rate. Additional references can be found in the Supplementary Data 2 . Extensive findings from our functional enrichment studies on these network driver causal genes, obtained from GO, KEGG, and Reactome, are presented in detail in Supplementary Data 6 . In Fig. 5 , we provide a detailed example of CREBBP , one of the genes identified by our framework and confirmed in the literature. This gene was chosen due to its higher number of connections compared to other genes, highlighting its essential role in the network. The plots of the other identified network driver protein-coding genes for Long COVID are provided in the Supplementary Data 7 . Download figure Open in new tab Fig. 5: Network plot highlighting a network driver gene for Long COVID. Our analysis identified CREBBP as a key network driver gene, supported by existing literature, with 273 total interactions (153 incoming, 120 outgoing). Connected genes are represented by three shapes based on network control properties: ellipses for critical genes (whose absence increases required driver nodes), diamonds for ordinary genes (whose removal maintains current driver nodes), and round rectangles for redundant genes (whose removal preserves network control). The three most enriched pathways are shown in green, purple, and blue, with node sizes proportional to their K-degree (network connectivity). 2.4 Gene Expression Clustering Reveals Long COVID Subtypes We clustered Long COVID patients into subgroups using gene expression data from the 32 identified causal genes. Moreover, we hypothesized that distinct gene expression patterns of risk/preventative genes and network driver genes underlay different clinical characteristics in patients. Using Consensus Clustering (ConC) [ 67 ], we identified patient subgroups that demonstrated coherent clustering and balanced distributions (i.e., not skewed toward a single subset). The analysis identified three distinct Long COVID subtypes, aligning with the three-cluster findings reported in previous research [ 4 , 45 ], each with high Average Silhouette Width (ASW) values indicating robust clustering: Cluster 1 included 65 individuals (ASW: 0.93), Cluster 2 contained 53 individuals (ASW: 0.85), and Cluster 3 consisted of 36 individuals (ASW: 0.75). As shown in Fig. 6 , these causal genes exhibited distinct expression patterns across subtypes, highlighting their potential role in differentiating symptom profiles. To explore this further, we mapped symptom prevalence within clusters to evaluate whether gene expression patterns reliably aligned with the identified symptoms. Significant differences in symptom distributions (p-value < 0.05) were observed, with symptoms grouped into broader categories, including respiratory, gastrointestinal, neurological, metabolic, psychological, dental, and sleep-related issues, enabling a comprehensive comparison across clusters. Table 7 complements this analysis by summarizing the key genes identified per cluster, their regulation patterns, biological functions, and associated enriched pathways contributing to the distinct Long COVID symptom manifestations [ 68 ]. Details of the RNA-seq and clinical datasets used in this analysis are provided in the Methods section. Download figure Open in new tab Fig. 6: Gene expression heatmap with hierarchical clustering of Long COVID patients revealing three distinct clusters. Normalized expression values range from -1 (dark purple, low expression) to 1 (bright yellow, high expression) across identified genes and patient clusters. Dendrograms on both axes show the hierarchical relationships between genes (top) and clusters (left), with height indicating the degree of similarity between clusters. Gene names are displayed on the x-axis, and cluster identifiers are shown on the y-axis. View this table: View inline View popup Download powerpoint Table 7: Gene expression patterns, pathways, and symptoms across Long COVID clusters. Cluster-specific genes highlight functions and enriched pathways associated with symptom persistence. The table shows relationships between clusters, symptoms, and pathways with significant biological relevance (p-values and FDR < 0.05). Clus.: Cluster. Reg.: Regulation. FDR: False Discovery Rate Cluster 1 showed a symptom profile dominated by respiratory issues and sleep disturbances. Increased mucus was reported by 29.23% of patients in this cluster, significantly higher than in cluster 2 (15.09%) and cluster 3 (16.67%) ( χ 2 p = 1.07 × 10 − 41 , adjusted p = 1.18 × 10 − 40 ). Lung problems (23.08%) and smell and/or taste problems (20.00%) were similarly more prevalent in cluster 1 ( χ 2 p = 7.69 × 10 − 7 , adjusted p = 1.21 × 10 − 6 ; χ 2 p = 8.23 × 10 − 15 , adjusted p = 2.59 × 10 − 14 , respectively). Sleep problems were also more common in cluster 1 (49.23%) compared to cluster 2 (28.30%) and cluster 3 (33.33%) ( χ 2 p = 9.98 × 10 − 33 , adjusted p = 5.49 × 10 − 32 ), aligning with previous reports indicating sleep disturbances as key features of specific Long COVID phenotypes [ 69 ]. This pattern is consistent with multiple cluster analyses identifying distinct respiratory and fatigue-related symptom groups [ 4 , 45 ]. The corresponding gene expression profile showed elevated expression of CREBBP , GRB2 , MAPK1 , and SMAD2 , which are involved in inflammatory responses, stress adaptation, and TGF-beta signaling pathways linked to respiratory function and sleep regulation. These molecular findings suggest that the selected genes in cluster 1 effectively capture the biological mechanisms underlying this group’s respiratory and sleep-related symptoms. A higher prevalence of psychological symptoms and dental issues characterized cluster 2. Anxiety and depression were observed in 37.74% of patients, slightly higher than in cluster 1 (36.92%) and significantly higher than in cluster 3 (25.00%) ( χ 2 p = 0.0082, adjusted p = 0.0106). Cavities and teeth problems affected 18.87% of cluster 2 patients, compared to 13.85% in cluster 1 and 5.56% in cluster 3 ( χ 2 p = 2.32 × 10 − 9 , adjusted p = 4.65 × 10 − 9 ). Gene expression analysis in cluster 2 revealed upregulation of CDC26 , CDKN1A , ESR1 , and YWHAG , genes associated with cell cycle regulation, stress response, and inflammation control, respectively. Notably, ESR1 has been implicated in psychiatric disorders, and YWHAG is known to modulate multiple signaling pathways relevant to mood regulation [ 70 ]. The prominence of neuropsychological symptoms in this cluster aligns with other Long COVID clustering studies that have identified distinct neurocognitive and mood-related phenotypes [ 4 , 45 ]. Furthermore, recent studies suggest an interplay between COVID-19 and oral health deterioration, providing a rationale for the increased dental issues in cluster 2 [ 71 ]. These findings reflect this cluster’s biological mechanisms tied to psychological and dental symptoms. Cluster 3 was defined by gastrointestinal (GI) symptoms and metabolic disturbances. A significant 38.89% of patients in this cluster experienced nausea, diarrhea, and/or vomiting, higher than in cluster 1 (13.85%) and cluster 2 (7.55%) ( χ 2 p = 1.37 × 10 − 116 , adjusted p = 3.02 × 10 − 115 ). Eating more or less was reported by 47.22% of patients in cluster 3, comparable to cluster 1 (47.69%) but higher than cluster 2 (37.74%) ( χ 2 p = 1.56 × 10 − 13 , adjusted p = 4.29 × 10 − 13 ). Headaches were also more common in cluster 3 (33.33%) compared to cluster 1 (30.77%) and cluster 2 (22.64%) ( χ 2 p = 2.19 × 10 − 20 , adjusted p = 8.02 × 10 − 20 ). The gene expression profile showed downregulation of HDAC1 , SRC , and TP53 , along with upregulation of NDUFA6 , genes associated with metabolic regulation, immune response, and cellular stress pathways. These alterations correlate with evidence of persistent metabolic and immune dysregulation in Long COVID [ 46 ]. The prominent GI issues are consistent with the recognition of GI-focused Long COVID clusters [ 4 , 45 ], showing the heterogeneous nature of post-COVID symptom profiles. These molecular profiles correlate with the GI and metabolic symptoms identified in cluster 3, highlighting the ability of these genes to capture the biological processes driving these manifestations. Integrating symptom profiles with gene expression clustering demonstrates how our identified genes stratify Long COVID patients into biologically distinct groups, each cluster exhibiting unique symptom signatures. Cluster 1 shows predominantly respiratory and sleep disturbances, suggesting potential benefits from therapies targeting respiratory and sleep pathways. Cluster 2 features psychological and dental issues, indicating the need for interventions addressing stress-related pathways and oral health. Cluster 3 presents GI and metabolic symptoms, suggesting treatments focused on metabolic and digestive support. The alignment between gene functions and symptom distributions validates the biological relevance of these causal genes and their role in driving diverse clinical manifestations. Further details, including complete statistical analyses and p-values, are available in Supplementary Data 8 . 3 Discussion and Conclusion Long COVID, or PASC, is characterized by the persistence or development of symptoms as a result of SARS-CoV-2 infection, affecting multiple systems, including respiratory, neurological, and cardiovascular systems [ 1 – 5 ]. Despite the increasing recognition of Long COVID, the causal genetic factors contributing to this condition remain primarily unidentified, posing a significant barrier to developing effective diagnostic and therapeutic strategies. Our work addressed this gap by introducing a novel multi-omics framework that integrated genetics, transcriptomics, and proteomics data to identify risk and preventive causal genes and network driver genes for Long COVID, ultimately guiding the development of targeted interventions. This approach enhances our understanding of Long COVID pathophysiology and provides a robust platform for further research into this disease. Previous studies have highlighted the complexity and heterogeneity of Long COVID, identifying various risk factors and biomarkers related to inflammation and immune function [ 7 ]. However, these observational studies have not established a causal relationship between genes and Long COVID. In contrast, our computational analysis identified several significant causal genes for Long COVID, supported by existing literature and novel candidates to direct future work. These genes are implicated in key biological processes potentially explaining the persistent symptoms observed in Long COVID cases. EA further highlighted pathways related to SARS-CoV-2 response, viral carcinogenesis, cell cycle regulation, and immune system processes, providing deeper insights into the pathophysiological mechanisms of Long COVID. Our analysis also demonstrated the utility of gene expression-based clustering in stratifying Long COVID patients into biologically meaningful subgroups, each with distinct symptom profiles and underlying molecular mechanisms. Cluster 1 exhibited respiratory and sleep disturbances linked to genes involved in inflammatory responses and TGF-beta signaling. Cluster 2 was characterized by psychological symptoms and dental issues associated with genes regulating stress response and cellular homeostasis. Cluster 3 showed GI and metabolic symptoms driven by genes related to metabolic regulation and digest pathways. These findings demonstrate the heterogeneity of Long COVID and highlight the potential for specific therapeutic interventions based on each cluster’s molecular and clinical characteristics. Future research should focus on validating these clusters and exploring personalized treatment strategies. Comparison with existing studies reveals that our integrated approach offers a comprehensive understanding of Long COVID’s genetic basis, providing robust causal evidence in the Long COVID network. Given that the genetics of Long COVID remain largely unknown, our framework uses the parameter α as an exploratory tool to discover potential causal genes from different analytical perspectives. We designed α as an adjustable parameter that allows researchers to investigate dynamically the interplay between MR and CT. To facilitate broader research and enable the discovery of new potential causal genes across different populations, we have open-sourced our framework and developed a user-friendly Shiny app ( https://sindypin.shinyapps.io/github/ ). This allows researchers to analyze their own datasets, explore different configurations, and reproduce our results. In particular, as α decreases, our model revealed potential network driver genes, highlighting factors such as TP53 , CREBBP , EP300 , YWHAG , SMAD3 , and GRB2 that may play crucial roles in network control. Our framework identified these candidate genes from the studied populations, providing initial insights into Long COVID genetics. This approach can be applied to emerging genetic studies, enabling both the validation of current findings and the discovery of new causal genes across diverse populations. This iterative discovery process will help build a more comprehensive understanding of the Long COVID’s genetic architecture. Our framework enables several future directions. As clinical and experimental studies establish the causal genes of Long COVID, researchers will be able to refine the α parameter for predictive applications. Additionally, integrating our approach with new data types, such as single-cell RNA-seq (scRNA-seq), could reveal more profound insights into cell-specific gene expression changes and heterogeneity. This could help identify how different cell types contribute to the development and persistence of Long COVID symptoms, potentially leading to more targeted therapeutic strategies. Our analysis of the causal genes’ involvement in other pathophysiological conditions revealed important connections that support and extend previous findings [ 7 , 47 ]. The overlap with conditions characterized by immune dysregulation, chronic inflammation, and multi-system involvement aligns with observed Long COVID manifestations [ 45 , 46 ]. Identifying genes involved in developmental disorders and inflammatory conditions is crucial, suggesting that genetic predisposition to certain disorders might influence Long COVID susceptibility [ 7 ]. This connection between other conditions and Long COVID provides additional context for understanding disease heterogeneity and could inform risk assessment strategies [ 47 ]. The database integration approach also highlights how existing knowledge of genetic disorders can inform our understanding of novel post-viral conditions, demonstrating the value of using established disease databases in new disease research. Beyond these immediate directions, the core set of 32 causal genes identified through our approach holds broad potential for clinical translation. Diagnostic kits using molecular assays (e.g., targeted polymerase chain reaction panels or multiplex protein-based tests) could be developed to detect abnormal expression patterns of these key genes, offering a more objective and standardized approach to diagnosing Long COVID. Integrating these biomarkers into machine learning models may further refine patient stratification, allowing reliable prediction of who is at higher risk and guiding early interventions. Such predictive models, trained on independent cohorts and incorporating multi-omic signatures, could ultimately be adapted into clinical decision support tools, helping healthcare professionals identify at-risk individuals before severe or persistent symptoms arise. In addition, these core causal genes may reveal novel therapeutic paths. Researchers can conduct targeted drug screening by identifying driver genes in the underlying molecular network or consider repurposing existing therapeutics. Interventions that modulate the activity or expression of these genes could potentially reduce Long COVID symptoms, complementing supportive care with more precise treatments. To realize these possibilities, further validation in in vitro and in vivo models and longitudinal studies are essential to assess whether modulating these genes influences patient outcomes over time. As data on cellular heterogeneity expands through scRNA-seq and other advanced profiling techniques, these causal genes can serve as molecular markers, guiding the search for cell-type-specific interventions and personalized medicine strategies. In conclusion, our study establishes a robust framework for identifying causal genes for Long COVID, providing critical insights into its pathophysiology and supporting the development of targeted therapeutic strategies. By integrating the discovered biomarkers into diagnostic assays, predictive modeling, and experimental therapeutics, we move closer to actionable clinical applications. Together, these findings have the potential to significantly improve the management and outcomes of individuals affected by Long COVID, offering a path toward more precise, evidence-driven care. 4 Methods 4.1 Input Data Collection and Preparation The success of our integrative multi-omics framework relies on the careful selection and preparation of diverse datasets that capture Long COVID’s genetic, transcriptomic, and proteomic dimensions. We collected and curated high-quality data from publicly available resources, ensuring robust coverage of key biological processes. These datasets include cis-eQTL information from the Genotype-Tissue Expression (GTEx) project [ 72 ], GWAS findings for Long COVID susceptibility [ 73 ], Whole Genome Sequencing (WGS) for Linkage Disequilibrium (LD) analysis [ 72 ], and gene-level data from Ensembl [ 74 ]. Additionally, RNA-seq [ 75 ], and PPI network datasets were incorporated to provide a comprehensive view of gene expression and functional interactions. The following subsections detail the sources, characteristics, and preparation steps for each dataset used in our analysis. 4.1.1 Expression Quantitative Trait Loci (eQTL) We utilized 49 significant cis-eQTL datasets, each within a 1Mb region and meeting an FDR threshold of < 0.05, obtained from the GTEx project (Version 8, Ensembl 99, GRCh38) [ 72 ]. These datasets contain 39,832 unique genes derived from nearly 1,000 healthy European individuals, accessed on 09 August 2023. They were crucial for investigating the relationship between genetic variation and gene expression across different human tissues ( Supplementary Data 9 ). For more details and a description of the datasets available in the GTEx consortium, refer to the original publication [ 76 ]. 4.1.2 Genome-Wide Association Studies (GWAS) We sourced a Long COVID GWAS dataset (Release 7; Ensembl 109; HGB GRCh38) from the study by Lammi et al., 2023 [ 73 ]. This dataset consists of 3,018 cases and 1,093,995 controls from six different ancestries, all of whom were evaluated for 19 symptoms three months post-COVID-19 infection according to the WHO and CDC definitions of Long COVID [ 1 , 2 ]. For comprehensive details, including the complete list of ancestries, symptoms, and unique SNPs, please refer to the detailed tables and figures in Supplementary Data 9 . 4.1.3 Whole Genome Sequencing (WGS) To ensure the robustness and validity of our method, we calculated the LD matrix retrieving the WGS BAM files from GTEx (Ensembl 88, GRCh38) that contains 820,792 unique SNPs from 836 European male and female individuals ( Supplementary Data 9 ). Access to this specific dataset was granted through special permission [ 72 ]. 4.1.4 Human Genes Dataset To assess each gene’s causal relationship with the outcome, we utilized the public Human Genes dataset from the Ensembl Genes database (version 110, GRCh38), which contains 70,116 genes [ 74 ]. 4.1.5 RNA Sequencing (RNA-seq) Moreover, we analyzed RNA-seq gene expression data from the Mount Sinai COVID-19 Biobank Study [ 68 ]. The dataset comprises patients with Long COVID symptoms (persisting for more than one-month post-acute infection, following established institutional criteria [ 3 – 5 ]), COVID-19 patients, and healthy controls. We sourced this dataset from the Gene Expression Omnibus - National Center for Biotechnology Information (GEO - NCBI) database under the identifier GSE215865, corresponding to the Ensembl GRCh37 release [ 75 ]. It contains 413 blood samples from 158 Long COVID individuals ( Supplementary Data 9 ). 4.1.6 Protein-Protein Interaction (PPI) Finally, we employed the human PPI dataset published by Vinayagam et al., 2011 [ 77 ] as a model for building the Long COVID network ( Supplementary Data 9 ). 4.2 Framework To create a comprehensive list of causal genes for Long COVID and understand their roles in regulating the disease, we used a fusion approach that integrated MR and CT. Specifically, we calculated the scores of each gene using the formula in Equation 1 . This approach produced a final ranking of genes based on their direct causal relationships and significance within the Long COVID network. The following sections detail the calculations of S Risk and S Network 4.2.1 Calculating S Risk To calculate S Risk , we employed the Mt-Robin method [ 10 ] to identify genes acting as risk or protective factors for Long COVID. Using GWAS ( D GWAS ) and eQTL ( D eQTL ) data ( Section 4.1 ), this approach accurately infers the dependence (fixed effects) between eQTL and GWAS effects, even with invalid IVs. We first constructed and refined the LD matrix using SNPs from our dataset to ensure robust genetic instruments. We calculated pairwise r 2 values and applied an LD 0.5 threshold to filter highly linked SNPs. Our multi-criteria SNP selection process eliminated those with multiple correlations above the LD threshold, prioritized SNPs present across multiple tissues with consistent effect directions, and selected significant SNPs with the smallest minimum p-values. Additionally, we required genes to be expressed in at least one tissue. The statistical analysis involved reverse regression coefficients and weighted regression with random slopes and correlated errors. We integrated these results with GWAS standard errors and the refined LD matrix to inform our resampling strategy. We evaluated causal relationships through bootstrapping by generating null distributions while preserving the SNP LD structure. We resampled GWAS effect sizes for each gene, maintaining LD correlations, and calculated test statistics under the null hypothesis of no association. The p-value for each gene was determined by the proportion of null test statistics exceeding the observed value, excluding samples with non-convergence or singular fits in the mixed-effects model. The final score calculation used the absolute effect size ( β y ) from the MR method. Genes with a p-value or FDR exceeding 0.05 received a score of 0, ensuring only significant causal effects. We normalized the MR score ( S Risk ) using min-max scaling for cross-gene comparability: where S MR represents each gene’s causal effect size, and min( S MR ) and max( S MR ) are the smallest and largest values across all genes, respectively. Finally, we estimated FDR-corrected p-values to identify statistically significant causal contributors to Long COVID (p-value < 0.05). 4.2.2 Calculating S Network To calculate S Network , we analyzed gene impact using Long COVID RNA-seq expression data and the human PPI network ( Section 4.1 ). Driver nodes are specific genes or proteins that, when manipulated, can control the state of the entire network. Network control refers to regulating or stabilizing the biological network by targeting these critical nodes. We first classified genes by removing each from the network and observing changes in the number of required driver nodes needed for control [ 13 ]. This process identified three categories: indispensable genes (increase in driver nodes needed), neutral genes (no significant change), and dispensable genes (minimal impact). We focused our analyses on indispensable genes due to their critical role in maintaining network control. We further refined indispensable genes into Type-I and Type-II classifications based on their network behavior. Type-I genes were categorized by their effect on other driver nodes. Critical genes were those whose removal increases the number of required driver nodes, particularly by disrupting directed paths that connect regulatory nodes to their downstream targets. Redundant genes decreased the number of required driver nodes, and ordinary genes did not change the number of required driver nodes. Type-II genes were classified based on their control requirements. Critical genes (zero in-degree, K in = 0) were present in all driver node sets, redundant genes were absent from all driver node sets, and ordinary genes were present in some, but not all driver node sets. We analyzed network connectivity using three measures: K (total degree), which represents total interactions and indicates network centrality; K in (in-degree), which shows incoming interactions that other genes could regulate; and K out (out-degree), which indicates outgoing interactions that influence different genes. The CT score ( S CT ) incorporated these classifications with weighted importance. Type-I critical genes received a weight of 1 as they are essential for network stability. Type-II critical genes received a weight of 2 as they must always be controlled (K in = 0). Redundant and ordinary genes received a weight of 0, reflecting their non-critical roles. We calculated S CT by multiplying each gene’s total degree (K) by its assigned weighted score (W): where S CT represents each gene’s network impact score calculated from its degree and weight. The final score was normalized using min-max scaling: where min( S CT ) and max( S CT ) are the smallest and largest values across all genes, respectively. 4.2.3 Analysis of Shared Genetic Basis Between Long COVID and Related Conditions Disease-gene associations were compiled using five complementary databases: MalaCards [ 78 ], DISEASES [ 79 ], DISGENET [ 80 ], MedGen [ 81 ], and GenCC [ 82 ]. We systematically queried these databases for conditions associated with our identified genes, focusing on pathophysiological features that overlapped with Long COVID manifestations. Selection criteria included: (1) presence of immune/inflammatory components, (2) chronic/persistent symptoms, (3) multi-system involvement, and (4) metabolic or endocrine disruption. Conditions were categorized based on their primary pathophysiological mechanisms and potential relevance to Long COVID pathogenesis. Database selection was based on comprehensive coverage of rare and common conditions, including mechanistic annotations and regular curation of disease-gene relationships. The complete dataset of conditions and their database sources is provided in Supplementary Data 4 . 4.2.4 Enrichment Analysis (EA) Our study conducted a comprehensive pathway EA on the risk, preventive, and network driver genes identified from our framework. The aim was to uncover the Biological Processes (BP), Cellular Components (CC), and Molecular Functions (MF) significantly associated with these genes. To ensure compatibility across various bioinformatics tools, we initially mapped gene IDs from Ensembl to Entrez ID using the org.Hs.eg.db database [ 74 ]. For the EA, we utilized well-established databases (GO [ 41 ], KEGG [ 42 ], and Reactome [ 83 ]). We prioritized enriched pathways based on statistical significance and relevance to established Long COVID literature. Pathways were considered significant when meeting all threshold criteria (p-value, p-adjust, and q-value < 0.05). Furthermore, we examined the Long COVID context by conducting a meaningful literature review to identify potential symptoms that may be related to each enriched pathway, providing additional insights into the disease’s possible clinical implications. We visualized the results using dot and network plots, representing enriched terms and molecular pathways clearly and intuitively. 4.2.5 Gene Expression Clustering We investigated Long COVID subtypes using gene expression data from our identified risk, preventive, and network driver genes. We determined the optimal number of clusters using the CancerSubtype package’s ConC algorithm [ 67 ], an unsupervised method for subtype discovery. The analysis used RNA-seq data detailed in Section 4.1.5 . Moreover, we performed a grid search across hyperparameters, evaluating 2 to 5 clusters with a fixed seed of 5 for reproducibility. After optimizing clustering parameters, we grouped Long COVID patients using the selected CC configuration. Cluster quality assessment involved calculating individual and group-wide silhouette widths. We selected the final number of clusters based on the highest ASW and balanced distribution of individuals across clusters. This clustering enabled the mapping of clinical data to analyze symptom prevalence within each subtype. To assess cluster-specific symptom patterns, we conducted statistical significance testing. We applied Chi-square tests when expected cell counts in contingency tables exceeded 5. We used Fisher’s exact test for cells with lower expected counts, simulated p-values (workspace: 2e8) for symptoms, and simulated Chi-square tests for other clinical variables. Statistical significance was set at p-value < 0.05. We then calculated symptom frequencies in absolute counts and relative percentages for each cluster, visualizing these distributions through comparative heatmaps. More details about the entire framework can be found in ( Supplementary Data 9 ) Data Availability All data produced in the present study are available upon reasonable request to the authors https://github.com/SindyPin/Causal-Multiomics-Method 5 Data Availability Tissue-specific eQTL datasets were sourced from the https://gtexportal.org/home/datasets (Version 8, Ensembl 99, GRCh38). Long COVID GWAS datasets were obtained from Lammi et al., 2023 [ 73 ]. The WGS data, essential for the LD matrix calculation, was retrieved from GTEx (Ensembl 88, GRCh38), which can be obtained following the guidelines specified at https://gtexportal.org/home/protectedDataAccess . The Human Genes dataset for the causal relationship assessment, was obtained from the Ensembl Genes database (version 110, GRCh38), available at https://www.ensembl.org/ . Long COVID RNA-seq data, used for critical gene discovery and molecular clustering, was derived from the Mount Sinai COVID-19 Biobank Study, available under GSE215865 at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE215865 . Finally, the PPI dataset, crucial for Long COVID network modeling, was retrieved from Vinayagam et al., 2011 [ 77 ]. 6 Code Availability The framework developed in this study to identify risk, preventive, and network driver genes is openly accessible for replication and further research at https://github.com/Causal-Multiomics-Method . This repository includes instructions for its use and a detailed guide to the data processing steps. Footnotes Contributing authors: sindy_licette.pinero{at}mymail.unisa.edu.au ; maisie.zhang{at}csiro.au ; lin.liu{at}unisa.edu.au ; Jiuyong.Li{at}unisa.edu.au ; hong.lee{at}unisa.edu.au ; marnie.winter{at}unisa.edu.au ; thin.nguyen{at}deakin.edu.au ; zjp{at}dali.edu.cn ; ↵ 1 Full link: https://sindypin.shinyapps.io/github/ References [1]. ↵ World Health Organization (WHO). Clinical case definition for Post COVID-19 condition ( 2021 ). URL https://www.who.int/teams/health-care-readiness/post-covid-19-condition . Accessed 14 Jan 2025 . [2]. ↵ Centers for Disease Control and Prevention (CDC). Long-term effects of COVID-19 ( 2023 ). URL https://www.cdc.gov/coronavirus/2019-ncov/long-term-effects/index.html . Accessed 14 Jan 2025 . [3]. ↵ Petersen , E. L. et al. Multi-organ assessment in mainly nonhospitalized individuals after SARS-CoV-2 infection: The Hamburg city health study COVID programme . European Heart Journal 43 , 1124 – 1137 ( 2022 ). OpenUrl PubMed [4]. ↵ Davis , H. E. , et al. Characterizing Long COVID in an international cohort: 7 months of symptoms and their impact . eClinicalMedicine ( 2021 ). [5]. ↵ Munblit , D. et al. Long COVID: Aiming for a consensus . Lancet Respiratory Medicine 10 , 632 – 634 ( 2022 ). OpenUrl PubMed [6]. ↵ Khullar , D. et al. Racial/ethnic disparities in Post-Acute Sequelae of SARS-CoV-2 infection in New York: an EHR-based cohort study from the RECOVER program . Journal of General Internal Medicine 38 , 1127 – 1136 ( 2023 ). OpenUrl PubMed [7]. ↵ Lai , Y. J. et al. Biomarkers in Long COVID-19: A systematic review . Frontiers in Medicine (Lausanne ) 10 ( 2023 ). [8]. ↵ Gasperi , C. , Chun , S. , Sunyaev , S. & Cotsapas , C . Shared associations identify causal relationships between gene expression and immune cell phenotypes . Communications Biology 4 ( 2020 ). [9]. ↵ Gaĺan , M. , et al. Persistent overactive cytotoxic immune response in a spanish cohort of individuals with Long-COVID: Identification of diagnostic biomarkers . Frontiers in Immunology 13 ( 2022 ). [10]. ↵ Gleason , K. J. , Yang , F. & Chen , L. S . A robust two-sample transcriptome-wide Mendelian Randomization method integrating GWAS with multi-tissue eQTL summary statistics . Genetic Epidemiology 45 , 353 – 371 ( 2021 ). OpenUrl PubMed [11]. ↵ Friedman , N. , Linial , M. , Nachman , I. & Pe’er , D . Using bayesian networks to analyze expression data . Journal of Computational Biology 7 , 601 – 620 ( 2000 ). OpenUrl CrossRef PubMed Web of Science [12]. ↵ Chaudhary , M. S. , Pham , V. V. H. & Le , T. D . NIBNA: a network-based node importance approach for identifying breast cancer drivers . Bioinformatics 37 , 2521 – 2528 ( 2021 ). OpenUrl CrossRef PubMed [13]. ↵ Vinayagam , A. et al. Controllability analysis of the directed human protein interaction network identifies disease genes and drug targets . Proceedings of the National Academy of Sciences of the United States of America 113 , 4976 – 4981 ( 2016 ). OpenUrl Abstract / FREE Full Text [14]. ↵ Pham , V. V. H. et al. CBNA: A control theory-based method for identifying coding and non-coding cancer drivers . PLOS Computational Biology 15 , e1007538 ( 2019 ). OpenUrl [15]. Bravaccini , S. et al. Estrogen and androgen receptor inhibitors: Unexpected allies in the fight against COVID-19 . Cell Transplantation 30 ( 2021 ). [16]. Ç etin , Z. , Bayrak , T. , Ŏgul , H. , Saygılı , E. & Akkol , E. Predicted SARS-CoV-2 miRNAs associated with epigenetic viral pathogenesis and the detection of new possible drugs for COVID-19 . Current Drug Delivery 18 , 1595 – 1610 ( 2021 ). OpenUrl PubMed [17]. ↵ Ni , J. , Qian , Y. & Zhang , Y . Identification and development of a five-gene signature to improve the prediction of mechanical ventilator-free days for patients with COVID-19 . European Review for Medical and Pharmacological Sciences 27 , 805 – 817 ( 2023 ). OpenUrl PubMed [18]. ↵ D’Agnillo , F. et al. Lung epithelial and endothelial damage, loss of tissue repair, inhibition of fibrinolysis, and cellular senescence in fatal COVID-19 . Science Translational Medicine 13 ( 2021 ). [19]. Villacampa , A. et al. SARS-CoV-2 s protein reduces cytoprotective defenses and promotes human endothelial cell senescence . Aging and Disease ( 2024 ). [20]. Temerozo , J. et al. VIP plasma levels associate with survival in severe COVID-19 patients, correlating with protective effects in SARS-CoV-2-infected cells . Journal of Leukocyte Biology 111 , 1107 – 1121 ( 2022 ). OpenUrl PubMed [21]. ↵ Wu , Y. et al. Sustained CREB phosphorylation by lipid-peptide liquid crystalline nanoassemblies . Communications Chemistry 6 , 241 ( 2023 ). OpenUrl PubMed [22]. ↵ Rehfeld , F. et al. CRISPR screening reveals a dependency on ribosome recycling for efficient SARS-CoV-2 programmed ribosomal frameshifting and viral replication . Cell Reports 42 , 112076 ( 2023 ). OpenUrl PubMed [23]. Vann , K. et al. Binding of the SARS-CoV-2 envelope e protein to human BRD4 is essential for infection . Structure 30 , 1224 – 1232 ( 2022 ). OpenUrl [24]. Iosef , C. et al. Plasma proteome of Long-COVID patients indicates HIF-mediated vasculo-proliferative disease with impact on brain and heart function . Journal of Translational Medicine 21 , 377 ( 2023 ). OpenUrl PubMed [25]. Herichová , I. , et al. Effect of 17beta-estradiol on the daily pattern of ACE2, ADAM17, TMPRSS2, and estradiol receptor transcription in the lungs and colon of male rats . PLoS One 17 ( 2022 ). [26]. Sonkar , C. et al. Repurposing of gastric cancer drugs against COVID-19 . Computers in Biology and Medicine 137 , 104826 ( 2021 ). OpenUrl PubMed [27]. Ren , J. et al. Possible pharmacological targets and mechanisms of sivelestat in protecting against acute lung injury . Computers in Biology and Medicine 170 , 108080 ( 2024 ). OpenUrl PubMed [28]. Aydemir , M. et al. Computationally predicted SARS-CoV-2 encoded microRNAs target NFKB , JAK/STAT, and TGFB signaling pathways. Gene Reports 22 , 101012 ( 2021 ). OpenUrl PubMed [29]. Trionfetti , F. et al. HDAC1-3 inhibition increases SARS-CoV-2 replication and productive infection in lung mesothelial and epithelial cells . Frontiers in Cellular and Infection Microbiology 13 , 1257683 ( 2023 ). OpenUrl [30]. Chen , H. et al. Analysing transcriptomic signatures and identifying potential genes for the protective effect of inactivated COVID-19 vaccines . PeerJ 11 ( 2023 ). [31]. ↵ Chu , Y. et al. Gene crosstalk between COVID-19 and preeclampsia revealed by blood transcriptome analysis . Frontiers in Immunology 14 , 1243450 ( 2024 ). OpenUrl PubMed [32]. Policard , M. , Jain , S. , Rego , S. & Dakshanamurthy , S . Immune characterization and profiles of SARS-CoV-2-infected patients reveal potential host therapeutic targets and SARS-CoV-2 oncogenesis mechanism . Virus Research 301 , 198464 ( 2021 ). OpenUrl PubMed [33]. ↵ Alpalhão , M. , Ferreira , J. & Filipe , P. Persistent SARS-CoV-2 infection and the risk for cancer . Medical Hypotheses 143 , 109882 ( 2020 ). OpenUrl PubMed [34]. Das , A. et al. Molecular and immune signatures, and pathological trajectories of fatal COVID-19 lungs defined by in situ spatial single-cell transcriptome analysis . Journal of Medical Virology 95 ( 2023 ). [35]. Zhang , L. , Zhu , K. , Xu , J. et al. Acetyltransferases CBP/p300 control transcriptional switch of beta-catenin and STAT1 promoting osteoblast differentiation . Journal of Bone and Mineral Research ( 2023 ). [36]. Meyer , B. et al. Characterising proteolysis during SARS-CoV-2 infection identifies viral cleavage sites and cellular targets with therapeutic potential . Nature Communications 12 , 5553 ( 2021 ). OpenUrl PubMed [37]. Norris , E. , Pan , X. & Hocking , D . Receptor-binding domain of SARS-CoV-2 is a functional alpha-V-integrin agonist . Journal of Biological Chemistry 299 , 102922 ( 2023 ). OpenUrl PubMed [38]. Major , J. et al. Type I and III interferons disrupt lung epithelial repair during recovery from viral infection . Science 369 , 712 – 717 ( 2020 ). OpenUrl Abstract / FREE Full Text [39]. Heydemann , L. et al. Hamster model for Post-COVID-19 alveolar regeneration offers an opportunity to understand Post-Acute Sequelae of SARS-CoV-2 . Nature Communications 14 , 3267 ( 2023 ). OpenUrl PubMed [40]. Vavougios , G . SARS-CoV-2 dysregulation of PTBP1 and YWHAE/Z gene expression: A primer of neurodegeneration . Medical Hypotheses 144 , 110212 ( 2020 ). OpenUrl PubMed [41]. ↵ The Gene Ontology Consortium . The Gene Ontology knowledgebase in 2023 . Genetics 224 ( 2023 ). [42]. ↵ Kanehisa , M. & Goto , S . KEGG: kyoto Encyclopedia of Genes and Genomes . Nucleic Acids Research 28 , 27 – 30 ( 2000 ). OpenUrl CrossRef PubMed Web of Science [43]. ↵ Yu , G. & He , Q.-Y . ReactomePA: An R/Bioconductor package for reactome pathway analysis and visualization . Molecular BioSystems 12 , 477 – 479 ( 2016 ). OpenUrl CrossRef PubMed [44]. ↵ Weisberg , E. et al. Repurposing of kinase inhibitors for treatment of COVID-19 . Pharmaceutical Research 37 , 167 ( 2020 ). OpenUrl CrossRef PubMed [45]. ↵ Niewolik , J. et al. Cluster analysis of Long COVID symptoms for deciphering a syndrome and its long-term consequence . Immunologic Research ( 2024 ). [46]. ↵ Wang , Z. & Gao , H . Anti-inflammatory or anti-SARS-CoV-2 ingredients in huashi baidu decoction and their corresponding targets: Target screening and molecular docking study . Arabian Journal of Chemistry 16 , 104663 ( 2023 ). OpenUrl PubMed [47]. ↵ Subramanian , A. et al. Symptoms and risk factors for Long COVID in nonhospitalized adults . Nature Medicine 28 , 1706 – 1714 ( 2022 ). OpenUrl CrossRef PubMed [48]. ↵ Boehm , J. W. , Lee , J. , Jones , D. & Freedman , D. E . Prevalence and risk factors for gastrointestinal symptoms after recovery from COVID-19 . Neurogastroenterology and Motility 34 ( 2022 ). [49]. ↵ Gay , L. , et al. vγ 9 vδ 2 t-cells are potent inhibitors of SARS-CoV-2 replication and represent effector phenotypes in patients with COVID-19 . Infectious Diseases 229 , 1759 – 1769 ( 2024 ). OpenUrl PubMed [50]. ↵ DePace , N. & Colombo , J . Long-COVID syndrome and the cardiovascular system: A review of neurocardiologic effects on multiple systems . Current Cardiology Reports 24 , 1711 – 1726 ( 2022 ). OpenUrl PubMed [51]. Messal , N. et al. Differential role for CD277 as a co-regulator of the immune signal in T and NK cells . European Journal of Immunology 41 , 3443 – 3454 ( 2011 ). OpenUrl CrossRef PubMed [52]. Loeffen , J. et al. cDNA of eight nuclear-encoded subunits of NADH:ubiquinone oxidoreductase: Human complex I cDNA characterization completed . Biochemical and Biophysical Research Communications 253 , 415 – 422 ( 1998 ). OpenUrl CrossRef PubMed Web of Science [53]. ↵ Gemble , S. et al. Pyrimidine pool disequilibrium induced by a cytidine deaminase deficiency inhibits PARP-1 activity, leading to the under replication of DNA . PLOS Genetics 11 ( 2015 ). [54]. ↵ Maas , S. , Gerber , A. & Rich , A . Identification and characterization of a human trna-specific adenosine deaminase related to the ADAR family of pre-mRNA editing enzymes . Proceedings of the National Academy of Sciences of the USA 96 , 8895 – 8900 ( 1999 ). OpenUrl Abstract / FREE Full Text [55]. ↵ Stiban , J. , Tidhar , R. & Futerman , A . Ceramide synthases: Roles in cell physiology and signaling . Advances in Experimental Medicine and Biology 688 , 60 – 71 ( 2010 ). OpenUrl CrossRef PubMed [56]. ↵ Frey , A. , Palenchar , D. , Wildemann , J. & Philpott , C . A glutaredoxin BOLA complex serves as an iron-sulfur cluster chaperone for the cytosolic cluster assembly machinery . Journal of Biological Chemistry 291 , 22344 – 22356 ( 2016 ). OpenUrl Abstract / FREE Full Text [57]. ↵ Sudre , C. H. et al. Attributes and predictors of Long COVID . Nature Medicine 27 , 626 – 631 ( 2021 ). OpenUrl CrossRef PubMed [58]. ↵ Zhou , Z. , He , M. , Shah , A. & Wan , Y . Insights into APC/C: From cellular function to diseases and therapeutics . Cell Div 11 , 9 ( 2016 ). OpenUrl CrossRef PubMed [59]. ↵ Del Rio , C. , Collins , L. F. & Malani , P . Long-term health consequences of COVID-19 . Journal of the American Medical Association 324 , 1723 – 1724 ( 2020 ). OpenUrl PubMed [60]. Baratchian , M. et al. Androgen regulation of pulmonary AR, TMPRSS2 and ACE2 with implications for sex-discordant COVID-19 outcomes . Scientific Reports 11 , 11130 ( 2021 ). OpenUrl PubMed [61]. Wu , C. et al. SARS-CoV-2 replication in airway epithelia requires motile cilia and microvillar reprogramming . Cell 186 , 112 – 130 ( 2023 ). OpenUrl CrossRef PubMed [62]. Liu , F. et al. Shared mechanisms and crosstalk of COVID-19 and osteoporosis via vitamin D . Scientific Reports 12 , 18147 ( 2022 ). OpenUrl PubMed [63]. Ripamonti , C. et al. HDAC inhibition as potential therapeutic strategy to restore the deregulated immune response in severe COVID-19 . Frontiers in Immunology 13 ( 2022 ). [64]. Cusato , J. et al. COVID-19: A possible contribution of the MAPK pathway . Biomedicines 11 , 1459 ( 2023 ). OpenUrl PubMed [65]. Lee , M. , Lee , S. & Bae , Y . Functional roles of sphingolipids in immunity and their implication in disease . Experimental and Molecular Medicine 55 , 1110 – 1130 ( 2023 ). OpenUrl CrossRef PubMed [66]. Basile , M. S. et al. The PI3K/AKT/mTOR pathway: A potential pharmacological target in COVID-19 . Drug Discovery Today 27 , 848 – 856 ( 2022 ). OpenUrl CrossRef PubMed [67]. ↵ Xu , T. & Le , T . Cancer subtypes identification, validation, and visualization based on gene expression data ( 2017 ). R package version 1.28.0. [68]. ↵ Thompson , R. et al. Molecular states during acute COVID-19 reveal distinct etiologies of long-term sequelae . Nature Medicine 29 , 236 – 246 ( 2023 ). OpenUrl CrossRef PubMed [69]. ↵ Crook , H. , Raza , S. , Nowell , J. , Young , M. & Edison , P . Long COVID—mechanisms, risk factors, and management . British Medical Journal 374 , 1648 ( 2021 ). OpenUrl CrossRef PubMed [70]. ↵ Hwang , W. J. , Lee , T. Y. , Kim , N. S. & Kwon , J. S . The role of estrogen receptors and their signaling across psychiatric disorders . International Journal of Molecular Sciences 22 , 373 ( 2020 ). OpenUrl PubMed [71]. ↵ Dickson-Swift , V. , Kangutkar , T. , Knevel , R. & Down , S . The impact of COVID-19 on individual oral health: A scoping review . BioMed Central Oral Health 22 , 422 ( 2022 ). OpenUrl PubMed [72]. ↵ GTEx portal - datasets ( 2023 ). URL https://gtexportal.org/home/datasets . Accessed 8 Sep 2023 . [73]. ↵ Vilma , L. et al. Genome-Wide Association Study of Long COVID . medRxiv ( 2023 ). [74]. ↵ Ensembl . Ensembl Genome Browser ( 2023 ). URL https://asia.ensembl.org/index.html . Accessed 5 Nov 2023 . [75]. ↵ NCBI GEO - GSE215865 ( 2023 ). URL https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE215865 . Accessed 11 Feb 2023 . [76]. ↵ GTEx Consortium . The GTEx consortium atlas of genetic regulatory effects across human tissues . Science 369 , 1318 – 1330 ( 2020 ). OpenUrl Abstract / FREE Full Text [77]. ↵ Vinayagam , A. et al. A directed protein interaction network for investigating intracellular signal transduction . Science Signal 4 ( 2011 ). [78]. ↵ MalaCards . The human disease database version 5.23 ( 2025 ). URL https://www.malacards.org . Accessed 13 Jan 2025 . [79]. ↵ DISEASES Database . Disease-gene associations ( 2025 ). URL https://diseases.jensenlab.org/Search . Accessed 13 Jan 2025 . [80]. ↵ DisGeNET . Integrating information on gene-disease associations ( 2025 ). URL https://www.disgenet.org . Accessed 13 Jan 2025 . [81]. ↵ MedGen . Medical genetics information ( 2025 ). URL https://www.ncbi.nlm.nih.gov/medgen . Accessed 13 Jan 2025 . [82]. ↵ GenCC . The gene curation coalition ( 2025 ). URL https://thegencc.org . Accessed 13 Jan 2025 . [83]. ↵ Gillespie , M. et al. The reactome pathway knowledgebase 2022 . Nucleic Acids Research ( 2021 ). View the discussion thread. Back to top Previous Next Posted February 12, 2025. Download PDF Data/Code Email Thank you for your interest in spreading the word about medRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Integrative Multi-Omics Framework for Causal Gene Discovery in Long COVID Message Subject (Your Name) has forwarded a page to you from medRxiv Message Body (Your Name) thought you would like to see this page from the medRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Integrative Multi-Omics Framework for Causal Gene Discovery in Long COVID Sindy Pinero , Xiaomei Li , Lin Liu , Jiuyong Li , Sang Hong Lee , Marnie Winter , Thin Nguyen , Junpeng Zhang , Thuc Duy Le medRxiv 2025.02.09.25321751; doi: https://doi.org/10.1101/2025.02.09.25321751 Share This Article: Copy Citation Tools Integrative Multi-Omics Framework for Causal Gene Discovery in Long COVID Sindy Pinero , Xiaomei Li , Lin Liu , Jiuyong Li , Sang Hong Lee , Marnie Winter , Thin Nguyen , Junpeng Zhang , Thuc Duy Le medRxiv 2025.02.09.25321751; doi: https://doi.org/10.1101/2025.02.09.25321751 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Genetic and Genomic Medicine Subject Areas All Articles Addiction Medicine (569) Allergy and Immunology (863) Anesthesia (300) Cardiovascular Medicine (4442) Dentistry and Oral Medicine (444) Dermatology (383) Emergency Medicine (609) Endocrinology (including Diabetes Mellitus and Metabolic Disease) (1511) Epidemiology (15230) Forensic Medicine (30) Gastroenterology (1126) Genetic and Genomic Medicine (6610) Geriatric Medicine (668) Health Economics (998) Health Informatics (4542) Health Policy (1370) Health Systems and Quality Improvement (1613) Hematology (543) HIV/AIDS (1266) Infectious Diseases (except HIV/AIDS) (15923) Intensive Care and Critical Care Medicine (1103) Medical Education (623) Medical Ethics (147) Nephrology (668) Neurology (6607) Nursing (346) Nutrition (999) Obstetrics and Gynecology (1146) Occupational and Environmental Health (957) Oncology (3338) Ophthalmology (974) Orthopedics (369) Otolaryngology (420) Pain Medicine (436) Palliative Medicine (130) Pathology (664) Pediatrics (1693) Pharmacology and Therapeutics (692) Primary Care Research (712) Psychiatry and Clinical Psychology (5448) Public and Global Health (9238) Radiology and Imaging (2202) Rehabilitation Medicine and Physical Therapy (1370) Respiratory Medicine (1196) Rheumatology (596) Sexual and Reproductive Health (714) Sports Medicine (530) Surgery (712) Toxicology (99) Transplantation (289) Urology (265) (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'a01db6061f111640',t:'MTc3OTgwODI3Mg=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();