Evidence-based genetic variants to gene mapping and prioritization uncovers distinct molecular pathophysiology and therapeutic landscape in polycystic ovary syndrome patients of different ethnicities.

Evidence-based genetic variants to gene mapping and prioritization uncovers distinct molecular pathophysiology and therapeutic landscape in polycystic ovary syndrome patients of different ethnicities. | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Evidence-based genetic variants to gene mapping and prioritization uncovers distinct molecular pathophysiology and therapeutic landscape in polycystic ovary syndrome patients of different ethnicities. Debojyoti De, Sindhuja Rajavelu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8610143/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 11 You are reading this latest preprint version Abstract Polycystic ovary syndrome (PCOS) is a highly prevalent and heterogeneous endocrine disorder affecting women of reproductive age, with substantial reproductive, metabolic, and long-term health consequences. While genome-wide association studies (GWAS) have identified multiple PCOS-associated loci across diverse populations, the functional interpretation of these predominantly non-coding variants and their translation into clinically actionable targets remain unresolved. Here, we present an integrative population-aware framework that systematically combines regulatory functional genomics, long-range chromatin interactions, genome-wide quantitative trait loci, and protein–protein interaction networks to prioritize effector genes underlying PCOS susceptibility. Applying this framework to East Asian and European populations, we demonstrate robust performance relative to existing approaches and uncover both shared and population-specific functions. Notably, our analyses reveal a predominant enrichment of metabolic dysregulation–associated pathways in East Asian PCOS, whereas European PCOS exhibits a stronger inflammatory and immune-related signature. These population-specific molecular phenotypes were further supported by transcriptomic data from PCOS patient samples. Importantly, integration of genetic evidence with network-based approach enabled the identification of druggable targets lacking direct genetic cues. Collectively, our study provides mechanistic insight into the ethnic heterogeneity of PCOS and establishes a scalable strategy for genetically informed, population-specific therapeutic prioritization, advancing precision medicine approaches for women’s health. Biological sciences/Computational biology and bioinformatics Health sciences/Diseases Biological sciences/Genetics Polycystic ovary syndrome variant to gene mapping epigenetic regulation clinical-proof-of-concept inflammation differential molecular pathology therapeutic landscape drug repositioning Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction PCOS is one of the most prevalent and complex endocrine disorders leading to anovulatory infertility among women of reproductive age. Beyond reproductive dysfunction, PCOS confers a lifelong increased risk for metabolic disorders including obesity, Type II Diabetes, dyslipidaemia, cardiovascular comorbidities, psychological distress and cancer 1 resulting in a substantial health and economic burden. Despite its high prevalence and multisystem involvement, effective disease-modifying management strategies for PCOS of this disorder is lacking, reflecting fundamental gaps in the understanding of its pathophysiology and aetiology. Currently, PCOS is primarily managed through organ-specific symptomatic therapies. Insulin resistance is commonly targeted using anti-diabetic agents, while reproductive and hyperandrogenic manifestations reproductive dysfunctions associated with PCOS are managed with combined oral contraceptives, ovulation inducers (Clomiphene citrate), aromatase inhibitors (letrozole) and antiandrogens (spironolactone) 2 . While these interventions provide partial symptomatic relief, they do not address the underlying molecular drivers of disease heterogeneity and fail to account for inter-individual and population-level variability in disease manifestation. This therapeutic stagnation highlights the urgent need for mechanistically informed approaches to manage PCOS. Accumulating evidence from familial and twin studies have revealed a strong genetic etiology behind PCOS suseptibility 3 , 4 . More recently, from genome-wide association studies and subsequent large-scale meta-analysis, several genetic variants spanning 16 loci in Han Chinese and European ethnicity 5 . Despite these findings, the exact mechanism by which these genetic variants confer disease phenotype is largely unknown. A major botteleneck arises from the observation that most of the PCOS-associated variants reside in non-coding regions of the genome, complicating their assignment to effector genes and obscuring the regulatory mechanisms through which they influence disease risk. Thus despite these recent advances in PCOS genomics, there remains no systematic framework that integrates regulatory genetics, population diversity, and druggability to inform precision therapeutics in women with PCOS. Importantly, genetic evidence has been shown to substantially increase the probability of success in drug development when robustly linked to target genes and disease mechanisms 6 – 9 . Bridging the gap between non-coding genetic variation and effector gene identification is therefore not only essential for elucidating PCOS biology, but also for nominating genetically informed therapeutic targets. Moreover, emerging GWAS data suggest that PCOS exhibits population-specific genetic architectures, raising the possibility that distinct molecular pathways may underlie disease manifestation across different ethnic groups. Thus, such population-specific differences may provide valuable biological insight into disease heterogeneity and inform population-aware strategies. In this study, we address these challenges by implementing an integrated, population-stratified (East Asian (EAS) European (EUR)) gene prioritization framework that systematically maps non-coding PCOS-associated variants to effector genes. We leverage complementary layers of regulatory functional genomics, including chromatin state annotations, long-range promoter–enhancer interactions derived from Hi-C studies, and genome-wide quantitative trait loci linking genetic variants to molecular traits (genome-wide QTL summary statistics). To further extend genetic evidence beyond direct variant–gene mappings, we integrate protein–protein interaction networks to identify functionally connected candidate genes lacking direct genetic signals. Integrating these two approaches and an effective scoring system we prioritized effector genes relevant to PCOS in East Asian and European populations. We demonstrated that although PCOS is characterized by common pathologies of hyperandrogenism, ovulatory dysfunction and polycystic ovarian morphology, there are subtle but biologically meaningful differences in the PCOS pathophysiological manifestation depending on ethnic background. Specifically, PCOS in East Asian populations is primarily characterized by enrichment of metabolic dysfunction–associated pathways, whereas European PCOS displays a stronger chronic inflammatory and immune-related signature. These findings were validated from transcriptomics data from PCOS patients too. Finally, we evaluate the druggability of the prioritized targets, highlighting potential therapeutic application and knowledge about pathways that may enable development of robust population-aware and genetically informed treatment strategies. Collectively, our findings provide a mechanistic framework for understanding ethnic heterogeneity in PCOS and also serves as an important advancement towards defining personalized therapeutic landscape. Results Differential prioritization of genetic targets in PCOS patients based on ethnic origin In order to identify, prioritize and define potential genetic targets and access their relevance in PCOS patients of different ethnic origin, we used the PCOS GWAS summary statistics from individuals of EAS and EUR populations (Table S1 , 1st Tab). Along with this, we integrated regulatory functional genomics data in the context of disease relevant tissues (Table S1 , 3rd Tab) and protein-protein interactions network using a previously reported prioritization pipeline, Priority Index algorithm as described in Fang et al. with a few modifications ( Materials and Methods , Fig. 1 and Figure S1 ) . Our prioritization process is a combination of multiple steps (Fig. 1 ) including (a) defining genomic predictors and scoring, (b) performance evaluation though benchmarking and validation of our pipeline and (c) defining the possible therapeutic landscape through target discovery (Fig. 1 ). Overview of the flow from variants to potential target discovery : The conceptual outline describes the flow of steps mainly categorized into 3 subparts : 1. Variant to gene mapping : GWAS summary statistics from EAS and EUR ethnicity and tissue-specific regulatory information were used to identify core genes through three main approaches: rGenes (physical proximity to SNPs and its regulatory state), qGenes (identified based on variant associating to some quantitative genomic traits ), and cGenes (based upon interaction of Enhancer-gene interaction facilitated by chromatin conformation change). These genes are weighted if they have further annotation evidences related to PCOS molecular functions (fGenes), related to PCOS phenotype (pGenes) and related to PCOS disease pathophysiology (dGenes). Additionally peripheral genes were identified using the knowledge of protein interaction space and connectivity to core genes. All genes were scored separately for each categories and these scores were finally combined. 2. Validation : Framework was benchmarked with respect to Open targets and further validated through enrichment analysis of PCOS drug targets in phase 2 and above in the scoring matrix. It was also validated by finding enrichment of differentially expressed genes of PCOS patients. 3. Analysis : Pathway-level integration identified genes mediating crosstalk between biological pathways, followed by cluster analysis to reveal therapeutic landscape of PCOS. In the process we could nominate several genes affiliating to different genomic and annotation predictors corresponding to all associated SNPs and scored them in different tissues. Defining gold standard positives (GSP) and negatives (GSN) for PCOS drug targets (Table S1 , 2nd Tab, see Materials and Methods) gave us opportunity to benchmark our approach. We combined the predictors using fishers combined meta-analysis (Table S2 , 1st Tab-S2, 14th Tab) which resulted in better performance in comparison to other similar methods (Figure S2 B (i) and S2C (i)). Further, we benchmarked our gene prioritization (Si approach) method with other competing approaches like text mining (curated literature evidences of gene associations to diseases) and genetic associations from open targets. We observed that our Si approach was highly comparable to the ‘text mining’ and significantly superior to ‘genetic association’ (Fig. 2 A (i) and 2B (i) for Ovary, and Figure S2 B (ii) and S2 C (ii) for other tissues) achieving an average AUC higher than 87% across tissues in both the population. Hence, we proceeded with this approach and ranked a total of ~ 17,500 genes across different tissues for both the population based on the integration of tissue-specific regulatory information, functional annotation and protein interaction networks (Fig. 1 ). The top ranked genes as predicted to be significant in the context of PCOS manifestation in ovary are presented in Manhattan plots, with the best 25 genes highlighted (Fig. 2 A (ii) and B (ii)). Among these, several genes are shared between population and many were population-specific. Interestingly, population specificity was more prominently reflected in the top 25 and further elaborated top 1% of the prioritized genes (Table S2 , 1st Tab-S2, 2nd Tab). Among the genes commonly prioritized in both the population in ovarian tissue, a few like ERBB3 (EAS:1; EUR:2), IKZF4 (EAS:2; EUR:25), RPS26 (EAS:4; EUR:4), EGFR (EAS:18; EUR:8), DENND1A (EAS:11; EUR:62) and YAP 1 (EAS:8; EUR:15) have been previously reported to be important in PCOS pathophysiology 10 – 12 . These genes ordered by significance, are broadly classified as core and peripheral genes, where cores are directly implicated by some variant association and thus affiliates to the genomic predictor. Whereas, the peripheral genes are those that are associated with the cores though network connectivity (Fig. 2 A (iii) and 2B (iii)) and do not have direct genomic evidences. It is noteworthy to mention that the peripheral genes that were retrieved in the process were not random and were reported to be contextual with respect to PCOS manifestation. For example, the core gene ERBB3 , present in both the population, interacts with the peripheral genes IGF1R, EGFR, GRB2 and CDH1 at the protein level in EAS (Fig. 2 A (iii)), a couple of which has been reported important in PCOS 12 , 13 . Also the same ERBB3 interacts with EGFR , CTNNB1 in EUR population (Fig. 2 B (iii)), both of which in turn interacts with NF-κB , together these candidates form a part of a chronic low-grade inflammatory axis 14 , 15 . Additionally EGFR, ERBB3 , and YAP1 present in the network of both the population, are involved in signaling by receptor tyrosine kinases and Hippo which are known for their role in PCOS 16 , 17 . Notably, within top 50 genes of the priority list (Table S2 ), genes like INS , INSR and IGF1R implicating insulin signaling and FSHR, LHCGR involving gonadotropin related signaling were exclusively present in the EAS population. On the other hand immune-related genes were differentially prioritized in the EUR population (Table S2 , 2nd Tab) with, NF-κB1, APP, IRF1 and few others (details in Table S2 , 2nd Tab). Incidentally, the gene APP, an inflammatory and oxidative stress associated gene, is a potential link to chronic inflammation in EUR women PCOS pathology. Next, to get a tissue-wise overview of the key prioritized genes as a reflection of the tissue-specific differential epigenetic regulation, we also ran the analysis to nominate cores and peripheral genes for other relevant tissues. We observed both tissue exclusive and common genes as depicted in the upset plot (Figure S3 A and Table S2 , 1st Tab-S2, 14th Tab). The common core genes showing high Si score in both the population includes RPS26 and ERBB3 (Figure S3 A (i) and (ii)), whereas the common peripheral genes (Figure S3 A (iii) and (iv)) across the population include EGFR and TP53 . Specifically for the EUR population, the commonly prioritized peripheral genes across tissues include the CTNNB1, APP, EP300 and EGFR which are all connected to the innate immune axis (Figure S3 B (ii) and Table S2 , 15th Tab) in some ways as mentioned above. Pathway enrichment identifies the importance of signal transduction and immune related pathways in polycystic ovary syndrome We next assessed the biology being manifested through the prioritized genes through enrichment studies using Reactome pathways. The top 25 pathways enriched in the top 100 prioritized genes in both the population in the ovarian tissue displayed a significant presence of various signaling, hormonal and immune related pathways (Fig. 2 A(iv) B(iv)). Interestingly, along with the core genes, many peripheral genes are also enriched in the relevant biological pathways. Specifically, signaling pathways involving EGFR , PDGFRA and FGFR which are exclusively enriched due to peripheral genes (Fig. 2 A (iv)) are within the top priority pathways in EAS population. These pathways are reported to be crucial in follicular growth and ovulation and reported to be dysregulated in PCOS 18 , 19 . Further, enrichment of IRAK1, IRAK and NF-κB signaling through peripheral genes in EUR population reflects the possible involvement of endosome-mediated innate immunity and inflammatory pathways in PCOS 20 . A list of all Reactome pathways enriched in top 100 prioritized genes of ovary across EAS and EUR populations are presented in Table S2 , 1st Tab-S2, 2nd Tab. Many top prioritized pathways in both the population like EGFR signalling in cancer and ERBB2 signalling (in EAS population), ERBB2 and p53 regulation pathways (in EUR population) (Figure S4 A) were consistently identified (Figure S4 B, Table S2 , 1st Tab-S2, 2nd Tab). Among the consistent pathways overlapping between the population ERBB2 signalling is the predominant one and key players involved are EGFR, RPS27A, UBA52, ERBB2, ERBB3 and ERBB4 (Figure S4 A). Apart from this, population-specific enrichment patterns were also evident. For example many growth factor related pathways were also constituent in EAS population. To systematically dissect the difference between the population specific pathway prioritization, we looked into three broad categories of pathways, Signal transduction, Immune response and Diseases (Figure S5 and S6). To assess the differences and similarities, we focussed on the top 100 statistically significant pathways enriched by the top 100 genes in EAS and EUR priority matrix. Both the population show prominent involvement of FGF and ERBB2 signalling with comparatively higher enrichment of this ERBB2 pathway in the EAS case (Figure S5 A and S6 A). Additionally, this population show more diverse set of protein tyrosine kinase and extensive growth factor related signalling pathways including Insulin receptor signalling cascade (Figure S5 A and C). In EUR, TAK1 dependent IKK and NF-κB signaling are suggestive towards the perturbation of TLR4-NF-κB-NLRP3 signaling axis 21 . Figure S6 B). Additionally, signalling by p75NTR and KIT are also among the prioritized pathways present (Figure S6 A and C). The immune related pathways are more pronounced in case of the EUR population Figure S5 B and S6 B). Clearly we observed a preferential prioritization in diverse immune pathways related to interleukins and cytokines in case of EUR population (Figure S6 ). Intrigued by the difference in immune related pathway enrichment and also because of few previous reports on associating PCOS with chronic low grade inflammation 15 , we set out to assess how different important immune related pathways (Reactome) are represented in ovary of both the population. When we looked into the list of top 100 genes in the priority matrix we found that the only a handful of important genes of key the immune pathways were present (Table S3 , 1st Tab). Most of the important immune related genes showed moderate to low ranks (Table S3 , 2nd Tab-S3, 6th Tab). For example, pro-inflammatory cytokines IL1B (EAS rank :2054 ; EUR rank :1567) IL18 (EAS rank: 4918; EUR rank : 3008) and TNF (EAS rank : 629; EUR rank : 503) which are important components of immune response and inflammation 22 , 23 (Table S3 , 4th Tab). The fact that important members of inflammasome and interleukins not highly rated through our gene prioritisation, implies that there are no direct genetic evidence to implicate their direct significance in PCOS pathophysiology. However, we observed that it is not them but their first degree interacting neighbours, which are present within top 100 gene list (Figure S7 A and Table S3 , 2nd Tab-S3, 9th Tab). For instance, a few of the inflammasome genes that exist within the top 100 ranked list are HSP90AB1 (EAS: Rank: 47, EUR: Rank: 77; in ovary matrix), APP (EAS: Rank: 91, EUR: Rank: 20; in ovary matrix), NF- κ B1 (EAS: ranked not within 100, EUR: ranked24; in ovary matrix) and RELA (EAS: ranked not within 100, EUR: Rank: 55; in ovary matrix). Again, NLRP3, NLRP1, and NLRC4 a key inflammasome marker, which poorly ranked in both the population, has its first neighbours all within top 100 in the list (Table S3 , 2nd Tab, Figure S7 B (i and ii)). With a similar trend, we further observed that there exists only 1out of 24 IL6 signalling genes in the top 100 list of EAS (p = 0.15 by Fisher’s exact test, Table S3 , 6th Tab). Whereas, 39 first degree neighbours of these 24 genes (out of 1205 in total, p = 7×10 − 19 , Table S3 , 6th Tab) are present in the top 100 rank. In alignment to this, we found that indeed, the ‘Si’ score distribution of members of both the inflammasome and IL6 signalling pathways depicts a very few overlap within top 100 genes (Figure S7 A), whereas there were many first degree neighbour overlap with the list. A network depiction further shows these scores differences for the putative inflammasome and their first degree neighbours (Figure S7 B i and ii) and as well as for IL6 signalling members and their first-degree neighbours (Figure S7 C i and ii). For cytokines and innate immune pathways, we did not find any preferential enrichment of first degree neighbours in top 100 genes with respect to the whole matrix in the EAS population (Table S3 , 1st Tab). To get a complete overview, we plotted the Si score distribution of members of several other immune, insulin receptor signalling (Table S3 , 7th Tab), GLUT4 translocation (Table S3 , 8th Tab) and androgen signalling (Table S3 , 9th Tab), all of which are well documented for their role in PCOS 24 . Clearly, we observed a very few members of these pathways were highly rated within top 100 and there is an increase in density of high Si score as we include their first degree neighbours (Figure S7 A) in all cases. Interestingly, we also observed that in case of EAS population most of the immune pathways probed here and GLUT4 translocation were considerably underrated with poor score density, in comparison to the EUR (Figure S7 A). In contrast, we found that insulin receptor signalling, genes related to glucose intolerance (Table S3 , 11th Tab) and insulin resistance phenotype (Table S3 , 10th Tab) had moderately higher score in EAS population (Figure S7 A and D). These observations were further validated when we got a similar trend of differential enrichment of these pathway genes in the relevant population. For instance, both the hormonal and the metabolic genes were enriched in EAS only, whereas the immune pathways demonstrated statistically significant enrichment only in EUR (Figure S7 E). Along with the ovary as the main tissue of interest for the disease manifestation, we also extensively studied the pathway enrichment and prioritization in other disease relevant tissues. We have already observed that the tissues shared a handful of common core and peripheral genes (Figure S3 A) and hence it was expected that there would be considerable overlap of pathways. Indeed, we observed pathways related to ERBB signaling, PI3K/AKT, Notch and cell cycle are commonly manifested in the relevant tissues of EAS population. Whereas, in case of EUR, along with all of the above mentioned pathways, there are prominent presence of pathways related to innate immunity (Figure S3 B and Table S2 ). EAS tissues display a noticeable manifestation of metabolic syndrome (Table S2 ). Tissue specific manifestation also include androgen biosynthesis pathways in adipose and pituitary and Leptin signaling in liver. Finally, we also created a consolidated network of the tissue specific pathways to understand how these pathways are shared between tissues (Figure S8 ) and to what extent they are exclusive. We observed that majority of the tissues are connected through the common pathways including ERBB-related signaling in both the population (Figure S8 A (i) and B (i)). Specifically, we observed that adipose-pituitary and cortex-liver tissue pairs share the maximum overlap of pathways in EAS and the EUR population respectively, an indication toward the shared biology. Moreover for both the populations, tissues form a closely-looped connection based on the number of overlapping pathways and we identified two such closely connected tissue groups: liver-cortex-pancreas and adipose-pituitary-ovary-muscle (Figure S8 A (ii) and B (ii)). Validating the method through recovery of clinically relevant targets: Next, to validate our approach in the context of the physiological relevance of PCOS, we sought to assess the capacity of this method to capture clinical proof of concept or the Gold standard positive (GSPs) drug targets of PCOS (in clinical trial phase II and above) during prioritisation. Target set enrichment analysis (TSEA) in ovary revealed 9 out of 21 (42%) and 12 out of 21 (57%) GSP drug targets were within the ‘leading edge’ for EAS (false discovery rate (FDR) = 7.1 × 10 − 4 ) and EUR (FDR = 6.5 ×10 − 4 ) respectively (Fig. 3 A (i) and 3B (i)). The leading-edge subset captures the important genes driving the enrichment signal. The most prominent among the recovered drug targets are FSHR (EAS: Rank: 6, EUR: Rank: 58; in ovary matrix), which is a Phase III drug target and ESR1 (EAS: Rank: 122, EUR: Rank: 95; in ovary matrix), which is an approved drug target targeted by Clomiphene 25 . Notably, ESR1 is a peripheral gene obtained through network connectivity. Another interesting target found within the leading edge is GLP1R targeted by Semaglutide, recently been investigated as a potent agent for weight loss 26 . Therapeutic Enrichment and Disease Relevance of Prioritized PCOS Genes Across population. (A, B) Target set enrichment analysis (TSEA) of 21 PCOS drug targets in ovary tissue using prioritized PCOS genes. (A (i) and B (i)) Leading edge plots show the distribution of known drug targets (Phase 2 and above) of PCOS within the ranked list of prioritized genes for ovarian tissue in the population as indicated. Enrichment scores are indicated by color represented by the Si score, and the red vertical line marks the leading edge (set of genes driving enrichment). (A (ii) and B (ii)) One-sided Fisher’s exact test quantifying enrichment of PCOS and other disease drug targets (Phase 2 and above) within the leading-edge genes. Comparisons are shown for core genes, core and peripheral genes combined together. (A (iii) and B (iii)) Enrichment of drug targets of other diseases within PCOS drug targets recovered (clinical-proof-of-concept targets) within the leading edge. (C) Gene Set Enrichment Analysis (GSEA) of Si-prioritized target genes in PCOS disease specific gene expression from various metabolic tissues from the CREEDS database. (i) EAS , (ii) EUR. Directionality (up/down) and study sources (GSE ID) are indicated. We then computed the enrichment of PCOS drug targets within the core genes as well as in core and peripheral genes combined, falling within the leading edge rank of the matrix and found them to be significantly enriched in both the population (Fig. 3 A (ii) and 3B (ii) left panels). Although, the enrichments were significant, the inclusion of peripheral genes did not show an improved enrichment for the PCOS targets. This could be because of less number of well characterised drug target for PCOS which are at higher phase of clinical trial. To understand further, we did the same exercise with drug targets for other diseases and found that the inclusion of peripheral genes through the knowledge of network connectivity indeed improved the enrichment (Fig. 3 A (ii) and 3B (ii) right panels). A few important diseases that got connected through these drug targets include obesity, type 2 diabetes, cardiovascular disorder and immune related ailments. This not only underscores the importance of network diffusion but also provides important clue about the potential repurposing through the approved targets of other diseases. Among the retrieved PCOS drug targets within the leading edge, 8 in EAS and 11 in EUR are also approved drug targets for other diseases (Fig. 3 A (iii), 3B (iii) and Table S4 , 3rd Tab). TSEA exercise for other relevant tissues additionally retrieved 10 or more PCOS drug targets, significantly enriched in the leading edge (Figure S9 (A) and (B)). Among these, AR, NR3C1, PGR, PPARG, GLP1R, LHCGR, FSHR, ESR1, ESR2 are commonly present within leading edge of all tissues in both the population (Table S4 , 1st Tab-S4, 2nd Tab). Some of these candidates are important drug targets of metabolic disorder. Interestingly, IL1R1 was seen in prefrontal cortex, adipose, pancreas and muscle in EUR and prefrontal cortex in EAS ethnicity. While, PNLIP is identified only in EAS liver and muscle. Therapeutic potential, as measured by combining coverage, FDR, NES, show significant recovery of GSPs (Figure S10 A (i) and (ii)) within pituitary and ovary showing maximum enrichment. Further, we validated our approach by probing enrichment of the DEGS from PCOS patients within the priority matrix (Fig. 3 C (i) and (ii)). Links between PCOS and a few metabolic disorder including obesity and type 2 diabetes (T2D) are well studied 27 and this prompted us to investigate enrichment of disease-specific DEGs (adipose, liver and pancreas) from obesity and T2D patients in the priority gene list of PCOS (Figure S10 B and C (i) and (ii)). While both the up and downregulated DEGs were well enriched in EAS population (NES up to 1.15 in obesity and NES up to 1.74 in T2D), in EUR primarily the upregulated DEGS were enriched. Pathway cross-talk identifies nodal points to perturb multiple pathways: In order to complement the individualistic prioritized biological pathways (as shown in Fig. 2 and Table S2 , 1st Tab-S2, 14th Tab), we set out to identify the interaction between pathways by merging the Reactome pathways on the background interactome (See methods). To this end, we constructed a maximum-scoring subgraph by maximising the number of highly prioritised genes within it along with a few low scoring interconnecting genes. Through this we obtained the crosstalk between pathways (Figure S11 and Table S2 , 1st Tab-S2, 14th Tab) in different tissues across population. With this exercise, we realized that the high ranking genes in the significance matrix, belonging to different pathways were all well connected with each other (Figure S12 A and B) validating the robustness of the network and significance of our prioritisation strategy. The few low scoring interconnecting genes include CALM1 (Rank 120, Ovary), CDK1 (Rank 159, Pancreas), CYCS (Rank 106, Cortex) in EAS and CYCS (Ranking above 160 in Liver, Muscle and Pituitary) in EUR. They are all important in linking the pathways and may represent important nodal points for perturbing multiple pathways together. Connecting these pathways through a minimal spanning tree, where the edges of connections are manifested by the number of shared genes between pathways (Fig. 4 A ii and B ii and Table S5 , 1st Tab), helped us to identify the common important nodal points of perturbation. We observed several genes like EGFR, KRAS , and NRAS , are shared between multiple signalling pathways including signalling by Tyrosine Kinase, Nuclear Receptor in EAS population (Fig. 4 A i, ii and iii). In case of EUR too, genes like RPS27A,TP53 , and UBA52 are shared between majority of the signalling including DNA repair, TAK1-dependent IKK and NF-κB activation, Regulation of TP53 activity (Fig. 4 B i, ii and iii), all of which have been implicated in impaired folliculogenesis and related pathophysiology 28 and could be well utilised to control multiple pathways simultaneously. The crosstalk genes and the pathways they are enriched in for other tissues are presented in Table S5 , 2nd Tab-S5, 7th Tab. Node removal from crosstalk network reveals resilience and identifies prospects of drug repurposing: Next, we further explored the contribution of the genes to the robustness and integrity of the crosstalk networks by accessing the tolerance of these network to node removal (Table S6 ), where we quantified the fraction of nodes disconnected after removal of a particular node (gene). It stems from an idea that a large number of nodes will be disconnected upon removal of a node critical for the network integrity. We observed that the crosstalk networks were quite tolerant (Table S6 , 1st Tab) to single node removal for both the population (Figure S12 A and B) with the maximum effect achieved through removal of TP53 (~ 13%) in case of EAS (Figure S12 A). Where as in case of EUR the effect of single node removal was not so prominent (Figure S12 B) with the highest of ~ 5% effect achieved upon removal of CDK9 . This robustness intrigued us to explore the effect of a combinatorial attack for node removal analysis to find out the smallest possible combination of nodes, ranging between two to four, mimicking a strategy very similar to multi target modulation to mitigate disease. We tried all combination including TP53 and CDK9 respectively for EAS and EUR, which would considerably destabilize the networks. We found that in EAS, TP53 in combination of GNAS achieved an effect of 23%, whereas, TP53 and GNAS when additionally combined with YWHAE , we observed a disconnection of 31% nodes (Figure S12 A). When we further performed the attack analysis with 4 nodes removal, we observed TP53, CTNNB1, RUNX1 and YWHAE increased effect to 46%. In EUR, two-node removal achieved 7.5%, and with the removal of three-node sets such as CDK9 - CDK1 - ELAVL1 or CDK9 - NEIL2-XRCC1 a disconnect of 12.5% was seen. Several four-node sets removal also achieved around 15% disconnect. All these observations were consistent with the idea that attacking multiple targets simultaneously are more effective to perturb resilient modules. Single-node removal in other tissues (Table S6 , 2nd Tab-S6, 7th Tab) identified tissue-specific vulnerabilities, with a highest impact for APP removal of 65.7% and 14.6% in adipose and liver respectively in EAS. In EUR however, single-node effects had low to modest effect (Table S6 , 2nd Tab-S6, 7th Tab). Node removal analysis further motivated us to evaluate the potential of perturbing the important cross talk genes through existing therapeutics. For the purpose we curated existing therapeutics from ChEMBL. We found considerable support in the potential to perturb a decent number of genes in both the crosstalk networks (EUR and EAS) from clinical evidences of phased and approved drug targets (approved drugs; FDR = 1.2×10 − 4 and phase 3 drugs for EAS, and approved drugs; FDR = 2.1×10 − 3 and phase 3 drugs for EUR) (Fig. 5 A i and B i respectively). This implies that all of these genes are important targets for other disease indications (Fig. 5 A iii and B iii). We further performed the attack analysis to identify specifically the contribution of these phased drug targets on the network integrity, as a measure of potential of drug repurposing. In alignment with the previous crosstalk found both the networks were robust to single node removal, showing a moderately low effect of attack. However, within the list, TP53 , a phase 3 drug target had the maximum effect of 13% disconnect (Fig. 5 A ii) in EAS. Additionally, in the context of this TP53 removal, combining CALM1 an approved drug target showed a decent 18% disconnect. CALM1 is approved drug target of cardiovascular disease highlighting the link between PCOS and CVD. On the other hand, the node removal analysis did not provide any strong contender that could have a considerable effect on network in case of EUR population (Fig. 5 B ii) with the only exception of CDK9 that have a marginal effect of 5% disconnect. Majority of the targets have similar effect on the network connection displaying a moderate but equal weightage to these drug targets. Network modularity study uncovers tissue-specific and shared cross-tissue functional modules: The study of pathway cross talks across tissues intrigued us to capture how different tissues are connected through these pathways across the population. To this end, we merged the pathway crosstalk instances from all the tissues together to obtain functional modules (based on network modularity) from the integrated network (Table S7 , 1st Tab and S7, 2nd Tab) and studied the prominence of different pathways in these modules across tissues. The exploration revealed six (M1-M6) modules in both EAS (Figure S13) and EUR (Figure S14) populations. Without any preconceived assumption and knowledge about functionality, we observed that each of these modules were linked to distinct molecular pathways underlying their association to non-random functional design (Table S7 ). We observed both shared and population-specific functional modules across tissues. Among the common pathways enriched by the modules in both the population, signalling by GPCRs, vesicle mediated transport, and translation are the most prominent ones (Table S7 and Figure S13 A and S14 A). Population specific modules are enriched in immune related pathways (M2) and DNA repair (M6) in EUR and various signalling axes (M6) and metabolism (M2) in EAS respectively. When we scored the modules in each of these populations with tissue-specific significance score, an interesting pattern emerged. We observed that the module containing genes related to several hormones ( TSHB, FSHB and CGB ) reportedly released from the pituitary 29 , 30 (M4 in both the population), specifically showed high intensity coloration when the modules were coloured by pituitary specific significance score (Figure S13 and S14 B). Moreover, in EAS ovary, highly-rated candidates including FSHR and LHCGR from M4 are connected to the majority of signalling genes in M6 via GNAS . A couple of these genes are also prominent members of Hypothalamus-pituitary axis 30 . Some other important genes specifically highlighted in pituitary only includes DRD1, DRD5, HTR6 and POMC genes, thus making the M4 truly exclusive for pituitary tissue. Signalling related to Vesicular Transport is enriched in module prioritized in muscle (Figure S13 and S14 B, Table S7 ), whereas the translation and RNA related metabolism are highlighted in the cortex. Interestingly, immune centric module M2 was specifically highlighted in liver and in alignment to this, a few important innate immunity genes containing module M6 ( IRF1, IFNA21, IFNA10, JAK, STAT3 etc.) was collectively prioritized in ovary, pancreas and adipose tissue in the EUR population. Collectively, these finding adds an additional level of validation for the relevance and significance of the scoring strategy and the pathway crosstalk. Self-organizing map identifies physiological similarity within tissues and defines therapeutic space: To understand more about the shared and unique cross talk genes between different tissues of both the population, we did a pairwise correlation analysis between tissue-specific significance scores from both the population (Fig. 6 A). We found weak or negative correlations between significance scores of tissues, except for a modest positive correlation for cortex and pituitary (Fig. 6 A). This underlines again a prominent distinction between the prioritized genes between the populations. To further understand this distinction and to capture the similarity between these tissues, we proceeded to find the genes that are shared between the tissues within a population. This population centric summary of the shared and unique genes participating in the crosstalk is important to define a common therapeutic space for multiple tissues involved in the disease. To this end, we used an unsupervised self-organising map approach to cluster the crosstalk genes in to a self-organized map (SOM) thereby depicting similar pattern of significance scoring in a cluster (Fig. 6 B (i) and C (i)). Tissue specific maps thus created placed tissues with similar profiles close to one another (Fig. 6 B (i) and 6C(i)). This revealed that there are both similarity and variability with respect to the significance scoring patterns of the cross-talk genes between tissues. Pituitary and adipose displayed high similarity in EAS in contrast to EUR population. Whereas, ovary and muscle had a higher similarity in clustering patterns in EUR when compared EAS. However the pancreas had a moderately similar distribution of the scoring pattern. These observation intrigued us to explore further to find out clusters in different tissues having similar scoring pattern and to create a consolidated supra-hexagonal cluster repressing all the tissues for a population. In the process, we identified 4 and 5 self-organising clusters in both EAS and EUR respectively (Figure S15) containing genes with similar pattern of significance scoring. The highly scored clusters emerged to be C3 and C2in EAS and EUR population respectively. These clusters highly scored genes in all the tissues analysed as displayed in the ridge plots with the score density mapping (Fig. 6 B (iv) and C (iv)). Interesting, for the EAS and EUR population, the same cluster happens to also contain the genes which are highly druggable as measured by the number of druggable pockets (Fig. 6 B (iii & v) and C (iii & v)). To get a biological overview of the clusters, we examined the pathways of these clusters. In this regard we performed pathway enrichment with the genes present in the druggable as well as high scored clusters (Figure S16 & Table S8 ). In both the population pathways were very similar as reported in Fig. 2 . Specifically, pathways related to signalling by receptor tyrosine kinase, nuclear receptors including Estrogen Receptor and signalling regulating membrane trafficking were prominently enriched in the high druggability and highly scored cluster ‘C3’ in the EAS (Figure S16A & Table S8 , 1st Tab). While, the Cluster C2 in EUR representing both druggable and well scored cluster, enriched in signalling related to tyrosine kinase, cell cycle and DNA repair (Figure S16B & Table S8 , 2nd Tab). Enrichment of pathways related to several immune related function was also in alignment with our previous observations (Fig. 2 ). Prioritized druggable clusters reveal drug repurposing prospects across populations: To explore the prospects of drug repurposing and also to assess the overall therapeutic landscape across tissues and population, we represented the cluster with high druggablility and significance scores in both the population as heat maps depicting their significance score (Figs. 7 and 8 ). The illustration also includes how these genes are affiliated to different reactome pathways. We observed that within the highly scored cluster C3 in EAS, the H29 displayed higher distribution of significance score for most of the genes in that hexagon across the tissues (Fig. 7 ). The genes in this hexagonal unit predominantly belongs to signalling by RTKs and Membrane trafficking. Few important genes in this hexagon include ERBB3, RAB5B, CDK2 etc. which are involved in pathways related to PCOS pathologies 31 , 32 . We found approved drugs targeting ERBB3 and RPS26 , which are prescribed for various cancers and muscular dystrophy respectively. The unit, H27 containing FSHR , a well-known candidate significantly involved in the biology pertaining to follicular development, also appeared as one of the promising targets. It is already in clinical investigation for PCOS being targeted by drugs Follitropin (phase 2) and Menotropin (phase3). The hexagonal unit H31 emerged as a critical hub with genes that are enriched in multiple pathways involving signalling by RTKs and nuclear receptors (ESR mediated signalling). This cluster includes approved drug targets and targets with high number of available druggable PDB structures including IGF1R, TP53, GRB2 and CTNNB1 . In the EUR population on the other hand (Fig. 8 ), the highly scored cum druggable cluster (cluster 2) is strongly associated with immune related pathways. The cluster H20, displayed high significance score across multiple tissues with approved drugs targets involved in of cell cycle, transcription and immune response pathways. The H37 containing highly scored IRF1 in majority of the metabolic tissues, is reported to be an important player in innate immune response. Similarly, in H8 presence of STAT1, NF- κ B1 and RELA are strongly associated with pathways related to immune response. All these observation signifies the potential of these highly scored clusters for further exploration with respect to drug discovery. Discussion PCOS affects around 10% to 13% women worldwide but still a lack of in depth understanding about the aetiology and the underlying mechanism of the disease manifestation, have restricted the development of an effective therapy of PCOS. However, some reports suggests that genetic, epigenetic and environmental factors contribute collectively towards the disease manifestation. Till date an extensive study towards the understanding of the interplay between the genetic and epigenetic regulation that might lead to the disease development and the pathophysiology, is lacking. Therefore, in this study we integrated PCOS-associated genetic variants, tissue-specific regulatory genomics and protein interaction network information to better understand the disease biology. Specifically, the inclusion of ChromHMM, genomic conformation and genome-wide QTL summary statistics (Fig. 1 ) to our study enabled us to leverage the regulatory annotations of the genomic variants. As majority of these data are tissue and population specific, it gave us a way to understand how disease biology is different due to different ethnic background and different disease relevant tissues. Scoring lead us to prioritize different genetic targets in EAS and EUR population. Inclusion of protein interaction information, resulted in prioritizing an additional list of secondary genetic targets, which happen to be closely associated with the primary genetic targets having genetic evidence, as in “guilt by association”. Indeed, for example in ovary, we noticed various important genes with respect to PCOS biology like IGF1R, INS, EGFR , ESR1, NF- κ B1, AMH and AR were picked up because of network connectivity. Eventually, many important signaling pathways were predominantly enriched due to these peripheral genes. Interestingly, differential scoring in different population revealed a preferential prioritization of immune related pathways in EUR. Whereas, metabolic and hormonal dysfunction are mostly predominant in the priority list of the EAS population with INSR, IGF1R, LHCGR, FSHR scoring high when compared to EUR prioritization. It is worthwhile to mention that PCOS has been linked to a systemic low grade inflammation 33 , 34 . To this end, we investigated the pattern of immune, metabolic and hormonal dysregulation in PCOS patients from a few gene expression dataset. In alignment to our prioritization we also observed a statistically significant preferential enrichment of immune related pathways in EUR and metabolic and hormonal pathways in EAS. This observation, on one hand underlined the difference in PCOS manifestation in different population, thereby indicating why we might need personalized strategies for PCOS management for different ethnicities. This observations also validated our scoring strategy. In fact, our strategy outperforms few other equivalent strategies of genetic target prioritization. Together, our findings highlight several key differences in etiology between populations. Among the important signaling in EAS population, signaling by receptor tyrosine kinases and insulin receptor are also reported elsewhere in association to PCOS. Additionally, prioritized genes in EAS ovary showed pronounced enrichment of multiple signal transduction pathways (Figure S5 ) including some relevant receptor tyrosine kinases (RTKs) (insulin-like growth factor 1 receptor and insulin receptor). All these suggest that EAS PCOS might be characterized by hyper activation of insulin/IGF-1-mediated metabolic and proliferative signaling cascades. The metabolic dysregulation extends beyond the ovary as evidenced by prioritization of insulin receptor signaling cascade in all the metabolic tissue and leptin signaling in liver 35 . The prominence of these pathways aligns with clinical observations that EAS women with PCOS exhibit more severe metabolic dysfunction and increased risk of metabolic complications compared to their EUR counterparts 36 – 38 . The enrichment of hormone ligand-binding receptors across adipose, liver, pituitary and ovary tissues, peptide hormone biosynthesis (adipose, liver and pituitary) and androgen biosynthesis (adipose, pituitary and ovary) suggest systemic hormonal dysfunction in EAS population. Similarly, the prioritization of the TLR4-NF-κB-NLRP3 signaling axis in EUR population, further offers a specific molecular hypothesis for this inflammatory phenotype unique to the population. This inflammatory signature may relate to the observation that EUR ancestry PCOS patients often present with systemic inflammation 39 , 40 . The systemic inflammatory profile is further evidenced by shared enrichment of IL21 and IL27 signaling across adipose, pancreas, and muscle, with IFNα and IFNβ signaling predominantly enriched in adipose tissue. The shared enrichment of ERBB2 signaling and PTK6 across both populations in most of the tissues suggests that certain core mechanisms is conserved regardless of ethnic background. We obtained an additional validation when using our priority list, we could retrieve a few potential PCOS drug targets that are being actively investigated in several stages of clinical trials. Enrichment of differentially expressed genes of obese and T2D patients in the prioritized gene list on the other hand, underscores the biological relevance of the PCOS associated co-morbidities. Our finding also establishes link though important genetic targets through which PCOS patients might get susceptible for many other metabolic disorders and cancer. Specifically, the role of several pathways involving protein tyrosine kinase, signaling by insulin receptors, inflammatory signaling etc. are implicated in both obesity and T2D 41 , 42 . All of these pathways were seen to be enriched in the context of PCOS also (Table S2 , 1st Tab-S2, 14th Tab). This observation warrants a further study towards the pleotropic effect of genetic variants which might lead to several associated co-morbidities including cardiovascular disease, hypertension, cancer, infertility and psychological disorders. In an attempt to explore more into the underlying mechanism of how different signaling pathways interact with each other, we identified the common genes through which these pathways cross talk. These common genes then become the hub to control multiple pathways together by targeting them by drugs. Identification of these target genes (nodes) and assessing the importance of these nodes through node removal analysis, gave us an idea of how some of these targets can be used for drug repurposing and to what extent these repurposing could be meaningful. From this exercise, we also realized that most of these genes are highly connected with each other and thus it is not so easy to perturb the disease network by targeting a single node. In some cases, combinatorial removal analysis was able to achieve a better effect, underlining the robustness of these networks, which eventually supports a polygenic model of disease manifestation 43 , where subtle expression changes of numerous target genes collectively shape disease manifestation. These findings also reinforce the notion that crosstalk between disease networks in complex traits are characterized by high interconnectivity and redundancy, making them intrinsically robust to perturbation. Pathway crosstalk also led us to explore cross tissue cross-talk which is an important step in understanding how the disease-relevant tissues collaborate with each other to manifest PCOS pathophysiology. We explored the modularity of the resulting integrated cross-tissue network and observed a few interesting modules enriched in specific tissues. Modules scoring high in cortex in both the population is enriched predominantly in cellular response to starvation and translation pathways. As it is reported that PCOS is often associated with reduced AMPK activity and parts of brain cortex and hypothalamus is responsible for starvation and hunger homeostasis, a connection to this pathway can be conceived as connection to metabolic imbalance and disruption of ovarian function 44 . The other important observation include the module enriched in pituitary specific TSHB and FSHB prioritization and high scoring of immune related genes in metabolic tissues in EUR, highlighting the involvement of these tissues in chronic inflammation 45 . Finally, the supra-hexagonal clustering provided an elegant, data-driven method to identify promising therapeutic targets by integrating tissue-wide similarity in scoring pattern and thus identifying potential therapeutic space and repurposing opportunities. For example, we could retrieve a few candidates like Follitropin and Menotropin targeting FSHR and Flutamide targeting AR (Figs. 7 and 8 ) which are under active clinical investigation. In a nutshell, our study focused on connecting the genetic variants to regulatory implications and thereby linking them to the pathophysiology of PCOS. Further, with differential prioritization of genetic targets in population and tissues we understood the subtle similarities and differences of the disease manifestation and thus opening up avenues to understand and treat the disease at a personalized scale. Thus our study redefines the ethnic heterogeneity in PCOS from being a mere confounding variable into a biologically informative dimension. Exploring how the disease related signaling propagates through various signaling axes and integrating tissue-wise patterns led us to define common therapeutic space, which needs to be explored further to develop new generation drugs for PCOS. Materials and Methods The pipeline: The overview of the entire flow has been illustrated in Fig. 1 . Population-specific GWAS summary statistics relevant to PCOS were systematically curated from publicly available database (see “ GWAS summary-level data collection and processing ”). To identify genes potentially influenced by PCOS-associated single nucleotide polymorphisms (SNPs), we employed three complementary approaches: (i) rGenes: Genes identified due to physical proximity to SNPs integrated with epigenetic regulatory information (ii) qGenes: Genes identified due to associations of the SNPs with a quantitative genomic trait (QTL) (iii) cGenes: Genes identified due to SNPs involved in chromatin interactions (see ‘ Identification of core genes under genomic influence ’). Together, these approaches identified a set of “core genes”, each supported by direct genomic evidence linking them to PCOS-associated SNPs. These genes were further annotated based on their relevance to PCOS, considering whether they had previously been implicated in the syndrome or associated with PCOS-related phenotypes (see ‘ Annotation of core genes with functional evidences ’). Next, we expanded the scope of gene identification by considering genes potentially influenced by network interactions. Using a curated Protein-Protein Interaction (PPI) network, we identified “peripheral genes”, subsequently both the core and the peripheral genes and were quantified based on their connectivity to each other using Random Walk algorithm (see ‘ Identification of peripheral genes with network evidence ’). The core and peripheral genes, along with their genomic and network evidences were combined into a gene-predictor matrix, where each gene was assigned a composite score ranging from 0 to 5 to reflect its overall clinical importance with respect to PCOS (see ‘ Gene-predictor matrix to gene prioritization ’). This comprehensive approach was applied specifically to PCOS-relevant tissues including the Ovary, Adipose, Liver, Muscle, Pancreas, Brain and Pituitary, ensuring that the prioritization process captured tissue-specific regulatory differences and population-level variation relevant to PCOS pathophysiology. To further investigate the functional roles of prioritized genes, we integrated pathway-level information into the PPI network, enabling the identification of genes that mediate interactions between distinct biological pathways (see ‘ pathway crosstalk ’). Next, we performed cluster analysis on the crosstalk-mediating genes across tissues to identify gene clusters having similarity in scoring pattern across tissues that may represent shared pathophysiology and define potential therapeutic space for PCOS (see ‘ Cluster analysis ’). The effectiveness of the prioritization framework was assessed using two performance evaluation strategies. We benchmarked the framework’s ability to distinguish GSPs from GSNs in a tissue-specific context. (See ‘ Benchmarking the scoring strategy ’). We examined the prioritization framework’s potential to retrieve the known therapeutics of PCOS (see ‘Target set enrichment analysis (Validation 1) and ‘Genetics-to-Current-Therapeutics (G2CT) potential’ ). A further validation was performed by gene set enrichment analysis of differentially expressed genes from PCOS patients. Detailed Materials and Methods are Declarations Author contributions S.R. collected the data, carried out the analyses according to the adopted pipeline. D.D. framed the research question, analyzed the data, supervised the work, and wrote the manuscript. All authors approved the final version of the manuscript. Funding There was no funding associated with this study. Disclosure statement No potential conflict of interest was reported by the author(s). Data availability statement Data sharing is not applicable to this study, as no new data were created. However, the authors confirm that the data supporting the findings of this study are available within the article, its supplementary materials, and referenced publications from which the data were extracted. References Stener-Victorin, E. et al. Polycystic ovary syndrome. Nat. Rev. Dis. Prim. 2024 101 10, 27- (2024). Akre, S., Sharma, K., Chakole, S. & Wanjari, M. B. Recent Advances in the Management of Polycystic Ovary Syndrome: A Review Article. Cureus 14, e27689 (2022). Vink, J. M., Sadrzadeh, S., Lambalk, C. B. & Boomsma, D. I. Heritability of polycystic ovary syndrome in a Dutch twin-family study. J. Clin. Endocrinol. Metab. 91, 2100–2104 (2006). Legro, R. S., Driscoll, D., Strauss, J. F., Fox, J. & Dunaif, A. Evidence for a genetic basis for hyperandrogenemia in polycystic ovary syndrome. Proc. Natl. Acad. Sci. U. S. A. 95, 14956 (1998). Day, F. et al. Large-scale genome-wide meta-analysis of polycystic ovary syndrome suggests shared genetic architecture for different diagnosis criteria. PLoS Genet. (2018) doi: 10.1371/journal.pgen.1007813 . Cook, D. et al. Lessons learned from the fate of AstraZeneca’s drug pipeline: a five-dimensional framework. Nat. Rev. Drug Discov. 13, 419–431 (2014). Nelson, M. R. et al. The support of human genetic evidence for approved drug indications. Nat. Genet. 47, 856–860 (2015). Waring, M. J. et al. An analysis of the attrition of drug candidates from four major pharmaceutical companies. Nat. Rev. Drug Discov. 14, 475–486 (2015). Uffelmann, E. et al. Genome-wide association studies. Nat. Rev. Methods Prim. 2021 11 1, 59- (2021). Ji, S. Y. et al. The polycystic ovary syndrome-associated gene Yap1 is regulated by gonadotropins and sex steroid hormones in hyperandrogenism-induced oligo-ovulation in mouse. Mol. Hum. Reprod. 23, 698–707 (2017). Censin, J. C., Bovijn, J., Holmes, M. V. & Lindgren, C. M. Colocalization analysis of polycystic ovary syndrome to identify potential disease-mediating genes and proteins. Eur. J. Hum. Genet. 29, 1446 (2021). Zhang, J. H. et al. Role of EGFR expressed on the granulosa cells in the pathogenesis of polycystic ovarian syndrome. Front. Endocrinol. (Lausanne). 13, 971564 (2022). Laban, M. A., Shemais, T. A., Elhelw, E. M. & El-Hassanin, A. R. Role Of Insulin Like Growth Factor in Polycystic Ovary Syndrome. Egypt. J. Hosp. Med. 72, 4005 (2018). Tan, W. et al. Insights on the NF-κB system in polycystic ovary syndrome, attractive therapeutic targets. Mol. Cell. Biochem. 479, 467–486 (2024). Rudnicka, E. et al. Chronic Low Grade Inflammation in Pathogenesis of PCOS. Int. J. Mol. Sci. 22, 3789 (2021). Bremer, A. A. & Miller, W. L. The serine phosphorylation hypothesis of polycystic ovary syndrome: a unifying mechanism for hyperandrogenemia and insulin resistance. Fertil. Steril. 89, 1039–1048 (2008). Clark, K. L. et al. Hippo Signaling in the Ovary: Emerging Roles in Development, Fertility, and Disease. Endocr. Rev. 43, 1074–1096 (2022). Haouzi, D. et al. Altered gene expression profile in cumulus cells of mature MII oocytes from patients with polycystic ovary syndrome. Hum. Reprod. 27, 3523–3530 (2012). Ziemak, H. et al. Angiogenesis in the ovary – Proangiogenic factors and their clinical applications. Biomed. Pharmacother. 191, 118416 (2025). Wang, J., Yin, T. & Liu, S. Dysregulation of immune response in PCOS organ system. Front. Immunol. 14, 1169232 (2023). Guo, Q. et al. NF-κB in biology and targeted therapy: new insights and translational implications. Signal Transduct. Target. Ther. 9, 1–37 (2024). Zhu, Q. & Kanneganti, T.-D. Distinct regulatory mechanisms control proinflammatory cytokines IL-18 and IL-1β. J. Immunol. 198, 4210 (2017). Fernandes, F. P., Leal, V. N. C., Souza de Lima, D., Reis, E. C. & Pontillo, A. Inflammasome genetics and complex diseases: a comprehensive review. Eur. J. Hum. Genet. 28, 1307–1321 (2020). Xu, Y. & Qiao, J. Association of Insulin Resistance and Elevated Androgen Levels with Polycystic Ovarian Syndrome (PCOS): A Review of Literature. J. Healthc. Eng. 2022, 9240569 (2022). Brown, J. & Farquhar, C. Clomiphene and other antioestrogens for ovulation induction in polycystic ovarian syndrome. Cochrane database Syst. Rev. 12, (2016). Austregésilo de Athayde De Hollanda Morais, B. et al. The efficacy and safety of GLP-1 agonists in PCOS women living with obesity in promoting weight loss and hormonal regulation: A meta-analysis of randomized controlled trials. J. Diabetes Complications 38, 108834 (2024). Kakoly, N. S. et al. Ethnicity, obesity and the prevalence of impaired glucose tolerance and type 2 diabetes in PCOS: a systematic review and meta-regression. Hum. Reprod. Update 24, 455–467 (2018). Xue, H. et al. The mechanism of NF-κB-TERT feedback regulation of granulosa cell apoptosis in PCOS rats. PLoS One 19, e0312115 (2024). McCartney, C. R., Campbell, R. E., Marshall, J. C. & Moenter, S. M. The role of gonadotropin-releasing hormone neurons in polycystic ovary syndrome. J. Neuroendocrinol. 34, e13093 (2022). Liao, B., Qiao, J. & Pang, Y. Central Regulation of PCOS: Abnormal Neuronal-Reproductive-Metabolic Circuits in PCOS Pathophysiology. Front. Endocrinol. (Lausanne) . 12, 667422 (2021). Padder, K. A., Yousuf, M. A., Jahan, N., Yousuf, S. D. & Ganie, M. A. Interplay between elevated RAB5B gene expression and insulin resistance among women with PCOS-insights from a case-control study. Endocrine 88, 323–329 (2025). Alan Harris, R. et al. Loci on chromosome 12q13.2 encompassing ERBB3, PA2G4 and RAB5B are associated with polycystic ovary syndrome. Gene 852, 147062 (2023). Escobar-Morreale, H. F., Luque-Ramírez, M. & González, F. Circulating inflammatory markers in polycystic ovary syndrome: A systematic review and metaanalysis. Fertil. Steril. 95, (2011). Aboeldalyl, S. et al. The Role of Chronic Inflammation in Polycystic Ovarian Syndrome—A Systematic Review and Meta-Analysis. Int. J. Mol. Sci. 22, 2734 (2021). Peng, Y. et al. Elevated Serum Leptin Levels as a Predictive Marker for Polycystic Ovary Syndrome. Front. Endocrinol. (Lausanne). 13, 845165 (2022). Guo, M. et al. Comparison of the phenotype of Chinese versus Dutch Caucasian women presenting with polycystic ovary syndrome and oligo/amenorrhoea. Hum. Reprod. 27, 1481–1488 (2012). Lo, J. C. et al. Epidemiology and Adverse Cardiovascular Risk Profile of Diagnosed Polycystic Ovary Syndrome. J. Clin. Endocrinol. Metab. 91, 1357–1363 (2006). Chahal, N. et al. Comparison of metabolic syndrome elements in White and Asian women with polycystic ovary syndrome: results of a regional, American cross-sectional study. F&S Reports 1, 305–313 (2020). Tian, D., Chen, J. & Liu, L. Causal relationship between inflammatory cytokines and polycystic ovary syndrome: a bidirectional mendelian randomization study. J. Ovarian Res. 17, 1–7 (2024). Rudnicka, E. et al. Inflammatory Markers in Women with Polycystic Ovary Syndrome. Biomed Res. Int. 2020, 4092470 (2020). Rohm, T. V., Meier, D. T., Olefsky, J. M. & Donath, M. Y. Inflammation in obesity, diabetes, and related disorders. Immunity 55, 31–55 (2022). Hotamisligil, G. S., Budavari, A., Murray, D. & Spiegelman, B. M. Reduced tyrosine kinase activity of the insulin receptor in obesity-diabetes. Central role of tumor necrosis factor-alpha. J. Clin. Invest. 94, 1543–1549 (1994). Boyle, E. A., Li, Y. I. & Pritchard, J. K. An expanded view of complex traits: from polygenic to omnigenic. Cell 169, 1177 (2017). Cao, Z. et al. Glycolytic Dysfunction in Granulosa Cells and Its Contribution to Metabolic Dysfunction in Polycystic Ovary Syndrome. Drug Des. Devel. Ther. 19, 5255 (2025). Rudnicka, E. et al. Inflammatory Markers in Women with Polycystic Ovary Syndrome. Biomed Res. Int. 2020, 4092470 (2020). Additional Declarations No competing interests reported. Supplementary Files TableS1.xlsx TableS2.xlsx TableS3.xlsx TableS4.xlsx TableS5.xlsx TableS6.xlsx TableS7.xlsx TableS8.xlsx Supplementarynotefinalversion.pdf Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 04 Apr, 2026 Reviews received at journal 03 Apr, 2026 Reviews received at journal 01 Apr, 2026 Reviewers agreed at journal 27 Mar, 2026 Reviewers agreed at journal 27 Mar, 2026 Reviewers agreed at journal 21 Jan, 2026 Reviewers agreed at journal 19 Jan, 2026 Reviewers invited by journal 19 Jan, 2026 Editor assigned by journal 19 Jan, 2026 Submission checks completed at journal 19 Jan, 2026 First submitted to journal 15 Jan, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8610143","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":577128241,"identity":"d29da866-4a4c-4613-ab75-af5535fdf8b8","order_by":0,"name":"Debojyoti De","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2ElEQVRIiWNgGAWjYDCCAwkMzAwMNgxsDAwGEBFmorQkpDGwsZGo5TADA1wLIcB3PPkAc+GP83l88s2bXxcw2MkzsPMewKtF8syzBOYZCbeL2djYyqxnMCQbNjDzJeDVYnAjx4CZJ+F2Yhsbj5kxDwMz0J08+F1ocCP/A1DLOZiWemK05ADVJBwAaTF+zMNwmLAWoF8MDs9ISwZqSStjnmFw3LCNkBZgiD18XGBjlzi/+fDmzwUV1fL8/GcIh/YBKM0mDYoaNoLqkQDzZ1JUj4JRMApGwcgBAM5JPFoZKn/SAAAAAElFTkSuQmCC","orcid":"","institution":"National Institute of Technology Durgapur","correspondingAuthor":true,"prefix":"","firstName":"Debojyoti","middleName":"","lastName":"De","suffix":""},{"id":577128242,"identity":"cd525a8e-a0fe-41ad-aa78-2dadaa24a0a2","order_by":1,"name":"Sindhuja Rajavelu","email":"","orcid":"","institution":"National Institute of Technology Durgapur","correspondingAuthor":false,"prefix":"","firstName":"Sindhuja","middleName":"","lastName":"Rajavelu","suffix":""}],"badges":[],"createdAt":"2026-01-15 11:53:36","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8610143/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8610143/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":100856444,"identity":"f5b9b6c5-cff8-4f7c-a9ad-568da5347da0","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25569377,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscriptwithmainfiguresnpjsba.docx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/abc87b2f55e5aea6f69416fe.docx"},{"id":100856435,"identity":"69821dcc-29e3-4d08-bd05-4738cc6e28a8","added_by":"auto","created_at":"2026-01-22 07:06:07","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4506,"visible":true,"origin":"","legend":"","description":"","filename":"98248e3645854adca9367310dbd783a3.json","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/e352304fdcafd5cbfd13c172.json"},{"id":100856439,"identity":"b9b3cb98-00aa-4224-b2d8-6c12ded74fda","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5589819,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarynotefinalversion.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/7e3e3ce69e1453c71538df97.pdf"},{"id":101202444,"identity":"0b2fd0f2-d4eb-4494-9a50-119ee9fa3e04","added_by":"auto","created_at":"2026-01-27 09:33:25","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":688385,"visible":true,"origin":"","legend":"","description":"","filename":"TableS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/5a48b4c57d032f541a464e89.xlsx"},{"id":100949343,"identity":"23883554-e4a5-41b9-b9f0-1bdd9587d1da","added_by":"auto","created_at":"2026-01-23 07:00:58","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1401430,"visible":true,"origin":"","legend":"","description":"","filename":"TableS2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/d604f98d2e6476953b473fa0.xlsx"},{"id":100856449,"identity":"c323ee02-01d8-4e73-bf3e-b46b8db46c31","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":872065,"visible":true,"origin":"","legend":"","description":"","filename":"TableS3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/1f0156a0e160924b7ea16854.xlsx"},{"id":100859445,"identity":"4a60b6e8-f89b-4e4b-acca-6cc98900dab7","added_by":"auto","created_at":"2026-01-22 07:28:09","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":194644,"visible":true,"origin":"","legend":"","description":"","filename":"TableS4.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/e57d1289ce5a71ad9e67f518.xlsx"},{"id":100859402,"identity":"4f749cc7-b448-426f-ac1b-04682870fc5e","added_by":"auto","created_at":"2026-01-22 07:27:43","extension":"xlsx","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":478343,"visible":true,"origin":"","legend":"","description":"","filename":"TableS5.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/97e23201d9a4de773777c2fb.xlsx"},{"id":100859401,"identity":"7895e21e-0051-4045-a624-06bb59ae3c86","added_by":"auto","created_at":"2026-01-22 07:27:43","extension":"xlsx","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":33675,"visible":true,"origin":"","legend":"","description":"","filename":"TableS6.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/69c5697a94c294395045c420.xlsx"},{"id":100856460,"identity":"9ff73b5a-b5b8-4c6a-80d5-4c41287e67b3","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"xlsx","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":27977,"visible":true,"origin":"","legend":"","description":"","filename":"TableS7.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/0706e434033cc9077940d2b1.xlsx"},{"id":100856447,"identity":"c1219e2c-4d5d-4752-b66c-225f53312601","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"xlsx","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53492,"visible":true,"origin":"","legend":"","description":"","filename":"TableS8.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/c2ab3cf5c905a724ec938668.xlsx"},{"id":100856455,"identity":"d461a7f2-8581-42b0-a003-2fd0c7b863e4","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"xml","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":139000,"visible":true,"origin":"","legend":"","description":"","filename":"98248e3645854adca9367310dbd783a31enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/b4ffaa1772bd477ecf394221.xml"},{"id":101296549,"identity":"982ef7f0-2afe-4814-a82f-5e328b41835b","added_by":"auto","created_at":"2026-01-28 09:14:27","extension":"jpeg","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1358610,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/42b0bbcdef987fd150b64ec3.jpeg"},{"id":100856459,"identity":"a01f230e-b089-46eb-b9da-30a0d3220e21","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7224929,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/813c6a3593bc649ee111799a.png"},{"id":100856454,"identity":"7ab862bb-c2f4-4013-b6b3-3c5039b4a9b8","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2249986,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/27e8255b7ea5ae4eaef32efa.png"},{"id":100856467,"identity":"4b36dadc-da31-480b-ba85-0405c4a596af","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3508683,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/a3d1bfd7192194b281eda0f8.png"},{"id":100856458,"identity":"4809da72-4c24-4e49-b655-ebd978b345b8","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3620125,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/1a787b326bde7b68e44b3d86.png"},{"id":100949581,"identity":"78a8b433-0ae7-4842-9a39-b155a0749d36","added_by":"auto","created_at":"2026-01-23 07:04:27","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3889778,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/b78a0f864afd903d26128823.png"},{"id":101296584,"identity":"922df08e-ab86-4188-8837-7463ea355b68","added_by":"auto","created_at":"2026-01-28 09:16:23","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1379725,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/6d7d803b8f80c6bc9209b0ff.png"},{"id":100856462,"identity":"ac5553b0-4d79-4e24-b79b-d58b19bb0a41","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1204327,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/8ed3407db321f6a186b819b4.png"},{"id":100856470,"identity":"982217b8-f2c2-45b2-9707-852cd3b895ed","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":266126,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/f095d26f9877a414e9663592.png"},{"id":100856468,"identity":"28fc1abb-7f81-4c9d-922b-53dc26ca43ab","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1125556,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/6e8f491bc6e5119db9bd4764.png"},{"id":100856464,"identity":"8d2a6638-36f3-4670-a9c5-70e433618c03","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":470500,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/f48a7981e7bb3e664dd0139c.png"},{"id":100856471,"identity":"0692247f-2e21-4517-a3c7-33d9ccab5371","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"png","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":553158,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/53b8fa99f502d7b4e1280f57.png"},{"id":100856472,"identity":"612618f8-d455-4326-a360-77ad45471b0b","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"png","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":519574,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/94459957b209e0172c22f3fe.png"},{"id":100859456,"identity":"2f5c4af4-c9d0-4f36-a4f5-9e3ba5194cdc","added_by":"auto","created_at":"2026-01-22 07:28:21","extension":"png","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":616790,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/0b6aef729e1364d752be9fc4.png"},{"id":100856453,"identity":"d8f45eab-e223-4001-a040-cbae41ccec1f","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"png","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":462068,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/e957696899322c740a69c0ca.png"},{"id":100856473,"identity":"40984a01-3c22-4a6a-a8d2-86b371c57814","added_by":"auto","created_at":"2026-01-22 07:06:10","extension":"png","order_by":27,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":386041,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/b44c14dfeb1cd08fa1a4fa8f.png"},{"id":100856466,"identity":"1446c976-aa49-41b3-9871-69dc68074ca5","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"xml","order_by":28,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":137493,"visible":true,"origin":"","legend":"","description":"","filename":"98248e3645854adca9367310dbd783a31structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/e3cde568fd73ca3bd1748490.xml"},{"id":100856469,"identity":"af1d380f-faef-4e50-b15b-5ba46f8dfb5d","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"html","order_by":29,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":154783,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/e061ced5bbf81006b64b9ab9.html"},{"id":100856433,"identity":"bde59a96-41d1-43f8-b615-c8e2164f887d","added_by":"auto","created_at":"2026-01-22 07:06:07","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":3439097,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eOverview of the flow from variants to potential target discovery: \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eThe conceptual outline describes the flow of steps mainly categorized into 3 subparts: \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e1.\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e \u003c/em\u003e\u003cem\u003e\u003cstrong\u003eVariant to gene mapping:\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e GWAS summary statistics from EAS and EUR ethnicity and tissue-specific regulatory information were used to identify core genes through three main approaches: rGenes (physical proximity to SNPs and its regulatory state), qGenes (identified based on variant associating to some quantitative genomic traits ), and cGenes (based upon interaction of Enhancer-gene interaction facilitated by chromatin conformation change). These genes are weighted if they have further annotation evidences related to PCOS molecular functions (fGenes), related to PCOS phenotype (pGenes) and related to PCOS disease pathophysiology (dGenes). Additionally peripheral genes were identified using the knowledge of protein interaction space and connectivity to core genes. All genes were scored separately for each categories and these scores were finally combined. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e2. Validation:\u003c/strong\u003e\u003c/em\u003e\u003cem\u003eFramework was benchmarked with respect to Open targets and further validated through enrichment analysis of PCOS drug targets in phase 2 and above in the scoring matrix. It was also validated by finding enrichment of differentially expressed genes of PCOS patients. 3. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003eAnalysis:\u003c/strong\u003e\u003c/em\u003e\u003cem\u003ePathway-level integration identified genes mediating crosstalk between biological pathways, followed by cluster analysis to reveal therapeutic landscape of PCOS.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/8c208ea5845faee58d695778.png"},{"id":100856436,"identity":"b5f1a2a0-5ed9-4ae5-b8ae-79478d20f36b","added_by":"auto","created_at":"2026-01-22 07:06:07","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":7224929,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eGene prioritization and benchmarking of the ovary: \u003c/strong\u003e\u003c/em\u003e\u003cem\u003e(A (i), B (i)) Benchmarking of the gene prioritization method used with respect to other similar method (Open Targets: Text Mining and Genetic Association), as measured by the area under ROC curves (AUC), depicting the capacity to separate the clinical-proof-of-concept targets from the simulated negative drug targets in EAS and EUR respectively. (A (ii), B (ii)) Genome-wide gene prioritization. Manhattan plots showing Significance Index (Si) scores across chromosomes, with top 25 genes labelled for the population as indicated. (A (iii), B (iii)) Protein-protein interaction subnetworks constructed with top-ranked genes in Ovary, colored by Si score, highlighting peripheral genes connected to core genes for both the population. (A (iv), B (iv)) Pathway enrichment analysis of the top 100 prioritized genes using one-sided Fisher’s exact test and significance derived from Benjamini-Hochberg method.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/61dd23a9a7df93170b0f2e1f.png"},{"id":100856437,"identity":"a338133d-3d8a-4567-85bf-9540513b15a3","added_by":"auto","created_at":"2026-01-22 07:06:07","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":2249986,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eTherapeutic Enrichment and Disease Relevance of Prioritized PCOS Genes Across population. (A, B)\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Target set enrichment analysis (TSEA) of 21 PCOS drug targets in ovary tissue using prioritized PCOS genes. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(A (i) and B (i)) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eLeading edge plots show the distribution of known drug targets (Phase 2 and above) of PCOS within the ranked list of prioritized genes for ovarian tissue in the population as indicated. Enrichment scores are indicated by color represented by the Si score, and the red vertical line marks the leading edge (set of genes driving enrichment). \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(A (ii) and B (ii))\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e One-sided Fisher’s exact test quantifying enrichment of PCOS and other disease drug targets (Phase 2 and above) within the leading-edge genes. Comparisons are shown for core genes, core and peripheral genes combined together. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(A (iii) and B (iii))\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Enrichment of drug targets of other diseases within PCOS drug targets recovered (clinical-proof-of-concept targets) within the leading edge. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(C)\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Gene Set Enrichment Analysis (GSEA) of Si-prioritized target genes in PCOS disease specific gene expression from various metabolic tissues from the CREEDS database. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(i)\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e EAS, \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(ii)\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e EUR. Directionality (up/down) and study sources (GSE ID) are indicated.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/0abf5c6c73315f1b4f773b40.png"},{"id":100856451,"identity":"bac38817-e4df-43cc-8679-1a95371ab409","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":3508683,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003ePathway Crosstalk. (A) EAS and (B) EUR\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Enrichment analysis of ovary pathway crosstalk genes for the populations as indicated illustrates (\u003c/em\u003e\u003cem\u003e\u003cstrong\u003ei)\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Top 10 enriched Reactome pathways identified by one-sided Fisher’s exact test and visualized as a kite plots and the genes present in those pathways as upset plots. Pathways are ranked by odds ratio and the size of the diamonds represents gene counts. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(ii)\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Minimum spanning tree of enriched crosstalk pathways, nodes representing pathways (sized by number of genes), edge thickness indicate overlapping genes between pathways, and node size corresponds to the number of genes from the cross talk list affiliated to those pathways. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(iii)\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Pathway interaction showing the genes (green triangles) contributing to connect pathways (red circles).\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/ca9cbed98e0891d60cd2cf9c.png"},{"id":100856456,"identity":"fb40ee34-d648-49d7-9b55-199eb4e16c26","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":3620125,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eDrug Repurposing Potential of PCOS Pathway Crosstalk Genes. (A) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eEAS \u003c/em\u003e\u003cem\u003e\u003cstrong\u003eand (B) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eEUR cohort. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(i) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eForest plots illustrating the enrichment of existing therapeutics for other disease indications within the pathway crosstalk Genes. Odds ratios is on x-axis and significance levels (–log₁₀ (FDR)) were computed using one-sided Fisher’s exact test. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(ii)\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Node (representing the drug targets of other diseases) removal analysis shows the fraction of nodes disconnected from the largest component upon single-node removal (y-axis). Genes with highest disruptive potential (sharp increases in disconnection) are indicated. The networks with these nodes indicated in color are shown in adjacent. The clinical phase information of these drug targets status are colored: pink for Phase 4 and purple for Phase 3 drugs. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(iii) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eBubble plots depict the drug targets (y-axis) mapped across various disease indications (x-axis). (Details in Table S6, 1\u003c/em\u003e\u003csup\u003e\u003cem\u003est\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e Tab).\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/4f4d1e5210123fdc1966b1c4.png"},{"id":100856461,"identity":"2a347102-9640-4c38-b50f-b90215c2104a","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":3889778,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003e(A) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eScatter plot representing the correlation between Significance scores for tissue-specific pathway crosstalk genes between in EAS (x-axis) and EUR (y-axis) population. The significance for the correlation calculated by Pearson’s test. Supra-hexagonal mapsrepresents the target clustering crosstalk genes across tissues on the basis of similarity in scoring pattern and druggability \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(B) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eEAS \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(C) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eEUR. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(i) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eComponent analysis of tissue-wise supra-hexagonal maps show prioritization similarity of crosstalk genes across tissues. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(ii) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eA druggability clustering presents the probability (of being a potential drug target) spread based on known or predicted druggable pockets. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(iii) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003ePolar pie charts represent the percentage of druggable genes within each cluster. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(iv) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eRidge plots show the distribution of Si score for genes in each cluster across tissues. Cluster C3 in EAS and C2 in EUR consistently exhibit high Si scores, indicating critical functional cluster. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003e(v and vi) \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eBubble plots show the druggable genes in the cluster (y-axis) and their PDB ids (x-axis). Color-coded is the number of druggable pockets predicted for their corresponding structures\u003c/em\u003e\u003cem\u003e\u003cstrong\u003e.\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/bdde866c0743f65800ad648c.png"},{"id":100859319,"identity":"7a63386f-6850-45a7-95a3-86771211d335","added_by":"auto","created_at":"2026-01-22 07:26:34","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":1379725,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eDrug repurposing evidence in highly scored clusters for EAS population. \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eThe heatmaps display target genes in cluster 3 across the relevant tissues. Rows represent target genes, and columns represent tissues. Annotations include: Top 25 enriched Reactome pathways associated with each gene, number of predicted druggable pockets based on protein structure, and approved drug indications with mechanisms of action targeting these genes in other diseases.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/537ee1be5bea01f752341230.png"},{"id":100856463,"identity":"e3c0e624-03ae-4b1f-8e7d-c3d044cbac9d","added_by":"auto","created_at":"2026-01-22 07:06:09","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":1204327,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eDrug repurposing evidence in highly scored clusters for EUR population. \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eTarget genes in cluster 2 across relevant tissues represented rows in heatmap and annotated for reactome pathways associated with each gene, containing druggable pockets, and approved drug target of other diseases in similar way as shown in Figure7.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image8.png","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/520346f43ea3e6dff31b2ea7.png"},{"id":101397684,"identity":"05a83b3c-daef-4ca8-bf5e-8dfb7fac4667","added_by":"auto","created_at":"2026-01-29 09:35:11","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":28076569,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/7ffadbf2-d64f-4892-8561-61387358db2f.pdf"},{"id":100859442,"identity":"0b08a76f-3f39-4026-b650-90028fb4f896","added_by":"auto","created_at":"2026-01-22 07:28:08","extension":"xlsx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":688385,"visible":true,"origin":"","legend":"","description":"","filename":"TableS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/da340f679da31ce173a410bb.xlsx"},{"id":100949354,"identity":"b514d281-dc04-4c60-bf6b-13e330f51901","added_by":"auto","created_at":"2026-01-23 07:01:04","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1401430,"visible":true,"origin":"","legend":"","description":"","filename":"TableS2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/0ed3694224217d8a6feae7be.xlsx"},{"id":100856441,"identity":"11ce9c18-1447-4e5b-9174-05ea63d8f098","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":872065,"visible":true,"origin":"","legend":"","description":"","filename":"TableS3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/4562491c78f278f5c6dad8d5.xlsx"},{"id":100856450,"identity":"ebb548b5-b01f-49b1-b42b-ea62d2d70684","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":194644,"visible":true,"origin":"","legend":"","description":"","filename":"TableS4.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/e5e9a713201bc19dc7947970.xlsx"},{"id":100859387,"identity":"9a54ca50-57da-4799-a31b-b2e33cf7fd5d","added_by":"auto","created_at":"2026-01-22 07:27:34","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":478343,"visible":true,"origin":"","legend":"","description":"","filename":"TableS5.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/7f0dc8bc863869fc5d88a106.xlsx"},{"id":100856440,"identity":"4ffd1ca2-cde8-4089-aabe-0ef8fd8f3680","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":33675,"visible":true,"origin":"","legend":"","description":"","filename":"TableS6.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/0dd0174eebb3fa1c3043ed6e.xlsx"},{"id":100856443,"identity":"4ac8f81d-c3d1-4064-a8a6-c9666c1eafd7","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":27977,"visible":true,"origin":"","legend":"","description":"","filename":"TableS7.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/50d96b6aa44308441715d47e.xlsx"},{"id":100856445,"identity":"d9c65f27-a573-41ec-a3f8-95659923038a","added_by":"auto","created_at":"2026-01-22 07:06:08","extension":"xlsx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":53492,"visible":true,"origin":"","legend":"","description":"","filename":"TableS8.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/1dd350b0a3e888c7c8ce1fd4.xlsx"},{"id":100859364,"identity":"98ce9f54-4b9a-4d73-94af-a960d972db3c","added_by":"auto","created_at":"2026-01-22 07:27:11","extension":"pdf","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":5589819,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarynotefinalversion.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8610143/v1/dcafe014498a8922455ddcf8.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Evidence-based genetic variants to gene mapping and prioritization uncovers distinct molecular pathophysiology and therapeutic landscape in polycystic ovary syndrome patients of different ethnicities. ","fulltext":[{"header":"Introduction","content":"\u003cp\u003ePCOS is one of the most prevalent and complex endocrine disorders leading to anovulatory infertility among women of reproductive age. Beyond reproductive dysfunction, PCOS confers a lifelong increased risk for metabolic disorders including obesity, Type II Diabetes, dyslipidaemia, cardiovascular comorbidities, psychological distress and cancer\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e resulting in a substantial health and economic burden. Despite its high prevalence and multisystem involvement, effective disease-modifying management strategies for PCOS of this disorder is lacking, reflecting fundamental gaps in the understanding of its pathophysiology and aetiology. Currently, PCOS is primarily managed through organ-specific symptomatic therapies. Insulin resistance is commonly targeted using anti-diabetic agents, while reproductive and hyperandrogenic manifestations reproductive dysfunctions associated with PCOS are managed with combined oral contraceptives, ovulation inducers (Clomiphene citrate), aromatase inhibitors (letrozole) and antiandrogens (spironolactone)\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. While these interventions provide partial symptomatic relief, they do not address the underlying molecular drivers of disease heterogeneity and fail to account for inter-individual and population-level variability in disease manifestation. This therapeutic stagnation highlights the urgent need for mechanistically informed approaches to manage PCOS.\u003c/p\u003e \u003cp\u003eAccumulating evidence from familial and twin studies have revealed a strong genetic etiology behind PCOS suseptibility\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. More recently, from genome-wide association studies and subsequent large-scale meta-analysis, several genetic variants spanning 16 loci in Han Chinese and European ethnicity\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Despite these findings, the exact mechanism by which these genetic variants confer disease phenotype is largely unknown. A major botteleneck arises from the observation that most of the PCOS-associated variants reside in non-coding regions of the genome, complicating their assignment to effector genes and obscuring the regulatory mechanisms through which they influence disease risk. Thus despite these recent advances in PCOS genomics, there remains no systematic framework that integrates regulatory genetics, population diversity, and druggability to inform precision therapeutics in women with PCOS. Importantly, genetic evidence has been shown to substantially increase the probability of success in drug development when robustly linked to target genes and disease mechanisms\u003csup\u003e\u003cspan additionalcitationids=\"CR7 CR8\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Bridging the gap between non-coding genetic variation and effector gene identification is therefore not only essential for elucidating PCOS biology, but also for nominating genetically informed therapeutic targets. Moreover, emerging GWAS data suggest that PCOS exhibits population-specific genetic architectures, raising the possibility that distinct molecular pathways may underlie disease manifestation across different ethnic groups. Thus, such population-specific differences may provide valuable biological insight into disease heterogeneity and inform population-aware strategies.\u003c/p\u003e \u003cp\u003eIn this study, we address these challenges by implementing an integrated, population-stratified (East Asian (EAS) European (EUR)) gene prioritization framework that systematically maps non-coding PCOS-associated variants to effector genes. We leverage complementary layers of regulatory functional genomics, including chromatin state annotations, long-range promoter\u0026ndash;enhancer interactions derived from Hi-C studies, and genome-wide quantitative trait loci linking genetic variants to molecular traits (genome-wide QTL summary statistics). To further extend genetic evidence beyond direct variant\u0026ndash;gene mappings, we integrate protein\u0026ndash;protein interaction networks to identify functionally connected candidate genes lacking direct genetic signals. Integrating these two approaches and an effective scoring system we prioritized effector genes relevant to PCOS in East Asian and European populations. We demonstrated that although PCOS is characterized by common pathologies of hyperandrogenism, ovulatory dysfunction and polycystic ovarian morphology, there are subtle but biologically meaningful differences in the PCOS pathophysiological manifestation depending on ethnic background. Specifically, PCOS in East Asian populations is primarily characterized by enrichment of metabolic dysfunction\u0026ndash;associated pathways, whereas European PCOS displays a stronger chronic inflammatory and immune-related signature. These findings were validated from transcriptomics data from PCOS patients too. Finally, we evaluate the druggability of the prioritized targets, highlighting potential therapeutic application and knowledge about pathways that may enable development of robust population-aware and genetically informed treatment strategies. Collectively, our findings provide a mechanistic framework for understanding ethnic heterogeneity in PCOS and also serves as an important advancement towards defining personalized therapeutic landscape.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eDifferential prioritization of genetic targets in PCOS patients based on ethnic origin\u003c/h2\u003e \u003cp\u003eIn order to identify, prioritize and define potential genetic targets and access their relevance in PCOS patients of different ethnic origin, we used the PCOS GWAS summary statistics from individuals of EAS and EUR populations (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e, 1st Tab). Along with this, we integrated regulatory functional genomics data in the context of disease relevant tissues (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e, 3rd Tab) and protein-protein interactions network using a previously reported prioritization pipeline, Priority Index algorithm as described in Fang et al. with a few modifications (\u003cb\u003eMaterials and Methods\u003c/b\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e\u003cb\u003e)\u003c/b\u003e. Our prioritization process is a combination of multiple steps (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) including (a) defining genomic predictors and scoring, (b) performance evaluation though benchmarking and validation of our pipeline and (c) defining the possible therapeutic landscape through target discovery (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eOverview of the flow from variants to potential target discovery\u003c/b\u003e: \u003cem\u003eThe conceptual outline describes the flow of steps mainly categorized into 3 subparts\u003c/em\u003e: \u003cb\u003e1. Variant to gene mapping\u003c/b\u003e: \u003cem\u003eGWAS summary statistics from EAS and EUR ethnicity and tissue-specific regulatory information were used to identify core genes through three main approaches: rGenes (physical proximity to SNPs and its regulatory state), qGenes (identified based on variant associating to some quantitative genomic traits ), and cGenes (based upon interaction of Enhancer-gene interaction facilitated by chromatin conformation change). These genes are weighted if they have further annotation evidences related to PCOS molecular functions (fGenes), related to PCOS phenotype (pGenes) and related to PCOS disease pathophysiology (dGenes). Additionally peripheral genes were identified using the knowledge of protein interaction space and connectivity to core genes. All genes were scored separately for each categories and these scores were finally combined.\u003c/em\u003e \u003cb\u003e2. Validation\u003c/b\u003e: \u003cem\u003eFramework was benchmarked with respect to Open targets and further validated through enrichment analysis of PCOS drug targets in phase 2 and above in the scoring matrix. It was also validated by finding enrichment of differentially expressed genes of PCOS patients. 3.\u003c/em\u003e \u003cb\u003eAnalysis\u003c/b\u003e: \u003cem\u003ePathway-level integration identified genes mediating crosstalk between biological pathways, followed by cluster analysis to reveal therapeutic landscape of PCOS.\u003c/em\u003e\u003c/p\u003e \u003cp\u003eIn the process we could nominate several genes affiliating to different genomic and annotation predictors corresponding to all associated SNPs and scored them in different tissues. Defining gold standard positives (GSP) and negatives (GSN) for PCOS drug targets (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e, 2nd Tab, see Materials and Methods) gave us opportunity to benchmark our approach. We combined the predictors using fishers combined meta-analysis (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e, 1st Tab-S2, 14th Tab) which resulted in better performance in comparison to other similar methods (Figure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003eB (i) and S2C (i)). Further, we benchmarked our gene prioritization (Si approach) method with other competing approaches like text mining (curated literature evidences of gene associations to diseases) and genetic associations from open targets. We observed that our Si approach was highly comparable to the \u0026lsquo;text mining\u0026rsquo; and significantly superior to \u0026lsquo;genetic association\u0026rsquo; (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA (i) and 2B (i) for Ovary, and Figure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e B (ii) and S2 C (ii) for other tissues) achieving an average AUC higher than 87% across tissues in both the population. Hence, we proceeded with this approach and ranked a total of ~\u0026thinsp;17,500 genes across different tissues for both the population based on the integration of tissue-specific regulatory information, functional annotation and protein interaction networks (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The top ranked genes as predicted to be significant in the context of PCOS manifestation in ovary are presented in Manhattan plots, with the best 25 genes highlighted (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA (ii) and B (ii)). Among these, several genes are shared between population and many were population-specific. Interestingly, population specificity was more prominently reflected in the top 25 and further elaborated top 1% of the prioritized genes (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e, 1st Tab-S2, 2nd Tab). Among the genes commonly prioritized in both the population in ovarian tissue, a few like \u003cem\u003eERBB3\u003c/em\u003e (EAS:1; EUR:2), \u003cem\u003eIKZF4\u003c/em\u003e (EAS:2; EUR:25), \u003cem\u003eRPS26\u003c/em\u003e (EAS:4; EUR:4), \u003cem\u003eEGFR\u003c/em\u003e (EAS:18; EUR:8), \u003cem\u003eDENND1A\u003c/em\u003e (EAS:11; EUR:62) and \u003cem\u003eYAP\u003c/em\u003e1 (EAS:8; EUR:15) have been previously reported to be important in PCOS pathophysiology\u003csup\u003e\u003cspan additionalcitationids=\"CR11\" citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThese genes ordered by significance, are broadly classified as core and peripheral genes, where cores are directly implicated by some variant association and thus affiliates to the genomic predictor. Whereas, the peripheral genes are those that are associated with the cores though network connectivity (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA (iii) and 2B (iii)) and do not have direct genomic evidences. It is noteworthy to mention that the peripheral genes that were retrieved in the process were not random and were reported to be contextual with respect to PCOS manifestation. For example, the core gene \u003cem\u003eERBB3\u003c/em\u003e, present in both the population, interacts with the peripheral genes \u003cem\u003eIGF1R, EGFR, GRB2\u003c/em\u003e and \u003cem\u003eCDH1\u003c/em\u003e at the protein level in EAS (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA (iii)), a couple of which has been reported important in PCOS\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Also the same \u003cem\u003eERBB3\u003c/em\u003e interacts with \u003cem\u003eEGFR\u003c/em\u003e, \u003cem\u003eCTNNB1\u003c/em\u003e in EUR population (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB (iii)), both of which in turn interacts with \u003cem\u003eNF-κB\u003c/em\u003e, together these candidates form a part of a chronic low-grade inflammatory axis\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Additionally \u003cem\u003eEGFR, ERBB3\u003c/em\u003e, and \u003cem\u003eYAP1\u003c/em\u003e present in the network of both the population, are involved in signaling by receptor tyrosine kinases and Hippo which are known for their role in PCOS\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Notably, within top 50 genes of the priority list (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e), genes like \u003cem\u003eINS\u003c/em\u003e, \u003cem\u003eINSR\u003c/em\u003e and \u003cem\u003eIGF1R\u003c/em\u003e implicating insulin signaling and \u003cem\u003eFSHR, LHCGR\u003c/em\u003e involving gonadotropin related signaling were exclusively present in the EAS population. On the other hand immune-related genes were differentially prioritized in the EUR population (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e, 2nd Tab) with, \u003cem\u003eNF-κB1, APP, IRF1\u003c/em\u003e and few others (details in Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e, 2nd Tab). Incidentally, the gene APP, an inflammatory and oxidative stress associated gene, is a potential link to chronic inflammation in EUR women PCOS pathology. Next, to get a tissue-wise overview of the key prioritized genes as a reflection of the tissue-specific differential epigenetic regulation, we also ran the analysis to nominate cores and peripheral genes for other relevant tissues. We observed both tissue exclusive and common genes as depicted in the upset plot (Figure \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e A and Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e, 1st Tab-S2, 14th Tab). The common core genes showing high Si score in both the population includes \u003cem\u003eRPS26\u003c/em\u003e and \u003cem\u003eERBB3\u003c/em\u003e (Figure \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e A (i) and (ii)), whereas the common peripheral genes (Figure \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e A (iii) and (iv)) across the population include \u003cem\u003eEGFR\u003c/em\u003e and \u003cem\u003eTP53\u003c/em\u003e. Specifically for the EUR population, the commonly prioritized peripheral genes across tissues include the \u003cem\u003eCTNNB1, APP, EP300\u003c/em\u003e and \u003cem\u003eEGFR\u003c/em\u003e which are all connected to the innate immune axis (Figure \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e B (ii) and Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e, 15th Tab) in some ways as mentioned above.\u003c/p\u003e \u003cp\u003e \u003cb\u003ePathway enrichment identifies the importance of signal transduction and immune related pathways in polycystic ovary syndrome\u003c/b\u003e \u003c/p\u003e \u003cp\u003eWe next assessed the biology being manifested through the prioritized genes through enrichment studies using Reactome pathways. The top 25 pathways enriched in the top 100 prioritized genes in both the population in the ovarian tissue displayed a significant presence of various signaling, hormonal and immune related pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA(iv) B(iv)). Interestingly, along with the core genes, many peripheral genes are also enriched in the relevant biological pathways. Specifically, signaling pathways involving \u003cem\u003eEGFR\u003c/em\u003e, \u003cem\u003ePDGFRA\u003c/em\u003e and \u003cem\u003eFGFR\u003c/em\u003e which are exclusively enriched due to peripheral genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA (iv)) are within the top priority pathways in EAS population. These pathways are reported to be crucial in follicular growth and ovulation and reported to be dysregulated in PCOS\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Further, enrichment of \u003cem\u003eIRAK1, IRAK\u003c/em\u003e and \u003cem\u003eNF-κB\u003c/em\u003e signaling through peripheral genes in EUR population reflects the possible involvement of endosome-mediated innate immunity and inflammatory pathways in PCOS\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. A list of all Reactome pathways enriched in top 100 prioritized genes of ovary across EAS and EUR populations are presented in Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e, 1st Tab-S2, 2nd Tab. Many top prioritized pathways in both the population like EGFR signalling in cancer and ERBB2 signalling (in EAS population), ERBB2 and p53 regulation pathways (in EUR population) (Figure \u003cspan refid=\"MOESM4\" class=\"InternalRef\"\u003eS4\u003c/span\u003e A) were consistently identified (Figure \u003cspan refid=\"MOESM4\" class=\"InternalRef\"\u003eS4\u003c/span\u003e B, Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e, 1st Tab-S2, 2nd Tab). Among the consistent pathways overlapping between the population ERBB2 signalling is the predominant one and key players involved are \u003cem\u003eEGFR, RPS27A, UBA52, ERBB2, ERBB3 and ERBB4\u003c/em\u003e (Figure \u003cspan refid=\"MOESM4\" class=\"InternalRef\"\u003eS4\u003c/span\u003e A). Apart from this, population-specific enrichment patterns were also evident. For example many growth factor related pathways were also constituent in EAS population. To systematically dissect the difference between the population specific pathway prioritization, we looked into three broad categories of pathways, Signal transduction, Immune response and Diseases (Figure \u003cspan refid=\"MOESM5\" class=\"InternalRef\"\u003eS5\u003c/span\u003e and S6). To assess the differences and similarities, we focussed on the top 100 statistically significant pathways enriched by the top 100 genes in EAS and EUR priority matrix. Both the population show prominent involvement of FGF and ERBB2 signalling with comparatively higher enrichment of this ERBB2 pathway in the EAS case (Figure \u003cspan refid=\"MOESM5\" class=\"InternalRef\"\u003eS5\u003c/span\u003e A and S6 A). Additionally, this population show more diverse set of protein tyrosine kinase and extensive growth factor related signalling pathways including Insulin receptor signalling cascade (Figure \u003cspan refid=\"MOESM5\" class=\"InternalRef\"\u003eS5\u003c/span\u003eA and C). In EUR, TAK1 dependent IKK and NF-κB signaling are suggestive towards the perturbation of TLR4-NF-κB-NLRP3 signaling axis\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Figure \u003cspan refid=\"MOESM6\" class=\"InternalRef\"\u003eS6\u003c/span\u003e B). Additionally, signalling by p75NTR and KIT are also among the prioritized pathways present (Figure \u003cspan refid=\"MOESM6\" class=\"InternalRef\"\u003eS6\u003c/span\u003e A and C). The immune related pathways are more pronounced in case of the EUR population Figure \u003cspan refid=\"MOESM5\" class=\"InternalRef\"\u003eS5\u003c/span\u003e B and S6 B). Clearly we observed a preferential prioritization in diverse immune pathways related to interleukins and cytokines in case of EUR population (Figure \u003cspan refid=\"MOESM6\" class=\"InternalRef\"\u003eS6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIntrigued by the difference in immune related pathway enrichment and also because of few previous reports on associating PCOS with chronic low grade inflammation\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e, we set out to assess how different important immune related pathways (Reactome) are represented in ovary of both the population. When we looked into the list of top 100 genes in the priority matrix we found that the only a handful of important genes of key the immune pathways were present (Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 1st Tab). Most of the important immune related genes showed moderate to low ranks (Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 2nd Tab-S3, 6th Tab). For example, pro-inflammatory cytokines \u003cem\u003eIL1B\u003c/em\u003e (EAS rank :2054 ; EUR rank :1567) \u003cem\u003eIL18\u003c/em\u003e (EAS rank: 4918; EUR rank : 3008) and \u003cem\u003eTNF\u003c/em\u003e (EAS rank : 629; EUR rank : 503) which are important components of immune response and inflammation \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e(Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 4th Tab). The fact that important members of inflammasome and interleukins not highly rated through our gene prioritisation, implies that there are no direct genetic evidence to implicate their direct significance in PCOS pathophysiology. However, we observed that it is not them but their first degree interacting neighbours, which are present within top 100 gene list (Figure \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e A and Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 2nd Tab-S3, 9th Tab). For instance, a few of the inflammasome genes that exist within the top 100 ranked list are \u003cem\u003eHSP90AB1\u003c/em\u003e (EAS: Rank: 47, EUR: Rank: 77; in ovary matrix), \u003cem\u003eAPP\u003c/em\u003e (EAS: Rank: 91, EUR: Rank: 20; in ovary matrix), \u003cem\u003eNF-\u003c/em\u003eκ\u003cem\u003eB1\u003c/em\u003e (EAS: ranked not within 100, EUR: ranked24; in ovary matrix) and \u003cem\u003eRELA\u003c/em\u003e (EAS: ranked not within 100, EUR: Rank: 55; in ovary matrix). Again, \u003cem\u003eNLRP3, NLRP1, and NLRC4\u003c/em\u003e a key inflammasome marker, which poorly ranked in both the population, has its first neighbours all within top 100 in the list (Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 2nd Tab, Figure \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e B (i and ii)). With a similar trend, we further observed that there exists only 1out of 24 IL6 signalling genes in the top 100 list of EAS (p\u0026thinsp;=\u0026thinsp;0.15 by Fisher\u0026rsquo;s exact test, Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 6th Tab). Whereas, 39 first degree neighbours of these 24 genes (out of 1205 in total, p\u0026thinsp;=\u0026thinsp;7\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;19\u003c/sup\u003e, Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 6th Tab) are present in the top 100 rank. In alignment to this, we found that indeed, the \u0026lsquo;Si\u0026rsquo; score distribution of members of both the inflammasome and IL6 signalling pathways depicts a very few overlap within top 100 genes (Figure \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e A), whereas there were many first degree neighbour overlap with the list. A network depiction further shows these scores differences for the putative inflammasome and their first degree neighbours (Figure \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e B i and ii) and as well as for IL6 signalling members and their first-degree neighbours (Figure \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e C i and ii). For cytokines and innate immune pathways, we did not find any preferential enrichment of first degree neighbours in top 100 genes with respect to the whole matrix in the EAS population (Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 1st Tab).\u003c/p\u003e \u003cp\u003eTo get a complete overview, we plotted the Si score distribution of members of several other immune, insulin receptor signalling (Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 7th Tab), GLUT4 translocation (Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 8th Tab) and androgen signalling (Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 9th Tab), all of which are well documented for their role in PCOS\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. Clearly, we observed a very few members of these pathways were highly rated within top 100 and there is an increase in density of high Si score as we include their first degree neighbours (Figure \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e A) in all cases. Interestingly, we also observed that in case of EAS population most of the immune pathways probed here and GLUT4 translocation were considerably underrated with poor score density, in comparison to the EUR (Figure \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e A). In contrast, we found that insulin receptor signalling, genes related to glucose intolerance (Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 11th Tab) and insulin resistance phenotype (Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e, 10th Tab) had moderately higher score in EAS population (Figure \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e A and D). These observations were further validated when we got a similar trend of differential enrichment of these pathway genes in the relevant population. For instance, both the hormonal and the metabolic genes were enriched in EAS only, whereas the immune pathways demonstrated statistically significant enrichment only in EUR (Figure \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e E).\u003c/p\u003e \u003cp\u003eAlong with the ovary as the main tissue of interest for the disease manifestation, we also extensively studied the pathway enrichment and prioritization in other disease relevant tissues. We have already observed that the tissues shared a handful of common core and peripheral genes (Figure \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e A) and hence it was expected that there would be considerable overlap of pathways. Indeed, we observed pathways related to ERBB signaling, PI3K/AKT, Notch and cell cycle are commonly manifested in the relevant tissues of EAS population. Whereas, in case of EUR, along with all of the above mentioned pathways, there are prominent presence of pathways related to innate immunity (Figure \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003eB and Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). EAS tissues display a noticeable manifestation of metabolic syndrome (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). Tissue specific manifestation also include androgen biosynthesis pathways in adipose and pituitary and Leptin signaling in liver. Finally, we also created a consolidated network of the tissue specific pathways to understand how these pathways are shared between tissues (Figure \u003cspan refid=\"MOESM8\" class=\"InternalRef\"\u003eS8\u003c/span\u003e) and to what extent they are exclusive. We observed that majority of the tissues are connected through the common pathways including ERBB-related signaling in both the population (Figure \u003cspan refid=\"MOESM8\" class=\"InternalRef\"\u003eS8\u003c/span\u003e A (i) and B (i)). Specifically, we observed that adipose-pituitary and cortex-liver tissue pairs share the maximum overlap of pathways in EAS and the EUR population respectively, an indication toward the shared biology. Moreover for both the populations, tissues form a closely-looped connection based on the number of overlapping pathways and we identified two such closely connected tissue groups: liver-cortex-pancreas and adipose-pituitary-ovary-muscle (Figure \u003cspan refid=\"MOESM8\" class=\"InternalRef\"\u003eS8\u003c/span\u003e A (ii) and B (ii)).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eValidating the method through recovery of clinically relevant targets:\u003c/h3\u003e\n\u003cp\u003eNext, to validate our approach in the context of the physiological relevance of PCOS, we sought to assess the capacity of this method to capture clinical proof of concept or the Gold standard positive (GSPs) drug targets of PCOS (in clinical trial phase II and above) during prioritisation. Target set enrichment analysis (TSEA) in ovary revealed 9 out of 21 (42%) and 12 out of 21 (57%) GSP drug targets were within the \u0026lsquo;leading edge\u0026rsquo; for EAS (false discovery rate (FDR)\u0026thinsp;=\u0026thinsp;7.1 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;4\u003c/sup\u003e) and EUR (FDR\u0026thinsp;=\u0026thinsp;6.5 \u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;4\u003c/sup\u003e) respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA (i) and 3B (i)). The leading-edge subset captures the important genes driving the enrichment signal. The most prominent among the recovered drug targets are \u003cem\u003eFSHR\u003c/em\u003e (EAS: Rank: 6, EUR: Rank: 58; in ovary matrix), which is a Phase III drug target and \u003cem\u003eESR1\u003c/em\u003e (EAS: Rank: 122, EUR: Rank: 95; in ovary matrix), which is an approved drug target targeted by Clomiphene\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Notably, \u003cem\u003eESR1\u003c/em\u003e is a peripheral gene obtained through network connectivity. Another interesting target found within the leading edge is \u003cem\u003eGLP1R\u003c/em\u003e targeted by Semaglutide, recently been investigated as a potent agent for weight loss\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eTherapeutic Enrichment and Disease Relevance of Prioritized PCOS Genes Across population. (A, B)\u003c/b\u003e \u003cem\u003eTarget set enrichment analysis (TSEA) of 21 PCOS drug targets in ovary tissue using prioritized PCOS genes.\u003c/em\u003e \u003cb\u003e(A (i) and B (i))\u003c/b\u003e \u003cem\u003eLeading edge plots show the distribution of known drug targets (Phase 2 and above) of PCOS within the ranked list of prioritized genes for ovarian tissue in the population as indicated. Enrichment scores are indicated by color represented by the Si score, and the red vertical line marks the leading edge (set of genes driving enrichment).\u003c/em\u003e \u003cb\u003e(A (ii) and B (ii))\u003c/b\u003e \u003cem\u003eOne-sided Fisher\u0026rsquo;s exact test quantifying enrichment of PCOS and other disease drug targets (Phase 2 and above) within the leading-edge genes. Comparisons are shown for core genes, core and peripheral genes combined together.\u003c/em\u003e \u003cb\u003e(A (iii) and B (iii))\u003c/b\u003e \u003cem\u003eEnrichment of drug targets of other diseases within PCOS drug targets recovered (clinical-proof-of-concept targets) within the leading edge.\u003c/em\u003e \u003cb\u003e(C)\u003c/b\u003e \u003cem\u003eGene Set Enrichment Analysis (GSEA) of Si-prioritized target genes in PCOS disease specific gene expression from various metabolic tissues from the CREEDS database.\u003c/em\u003e \u003cb\u003e(i)\u003c/b\u003e \u003cem\u003eEAS\u003c/em\u003e, \u003cb\u003e(ii)\u003c/b\u003e \u003cem\u003eEUR. Directionality (up/down) and study sources (GSE ID) are indicated.\u003c/em\u003e\u003c/p\u003e \u003cp\u003eWe then computed the enrichment of PCOS drug targets within the core genes as well as in core and peripheral genes combined, falling within the leading edge rank of the matrix and found them to be significantly enriched in both the population (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA (ii) and 3B (ii) left panels). Although, the enrichments were significant, the inclusion of peripheral genes did not show an improved enrichment for the PCOS targets. This could be because of less number of well characterised drug target for PCOS which are at higher phase of clinical trial. To understand further, we did the same exercise with drug targets for other diseases and found that the inclusion of peripheral genes through the knowledge of network connectivity indeed improved the enrichment (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA (ii) and 3B (ii) right panels). A few important diseases that got connected through these drug targets include obesity, type 2 diabetes, cardiovascular disorder and immune related ailments. This not only underscores the importance of network diffusion but also provides important clue about the potential repurposing through the approved targets of other diseases. Among the retrieved PCOS drug targets within the leading edge, 8 in EAS and 11 in EUR are also approved drug targets for other diseases (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA (iii), 3B (iii) and Table \u003cspan refid=\"MOESM4\" class=\"InternalRef\"\u003eS4\u003c/span\u003e, 3rd Tab). TSEA exercise for other relevant tissues additionally retrieved 10 or more PCOS drug targets, significantly enriched in the leading edge (Figure \u003cspan refid=\"MOESM9\" class=\"InternalRef\"\u003eS9\u003c/span\u003e (A) and (B)). Among these, \u003cem\u003eAR, NR3C1, PGR, PPARG, GLP1R, LHCGR, FSHR, ESR1, ESR2\u003c/em\u003e are commonly present within leading edge of all tissues in both the population (Table \u003cspan refid=\"MOESM4\" class=\"InternalRef\"\u003eS4\u003c/span\u003e, 1st Tab-S4, 2nd Tab). Some of these candidates are important drug targets of metabolic disorder. Interestingly, \u003cem\u003eIL1R1\u003c/em\u003e was seen in prefrontal cortex, adipose, pancreas and muscle in EUR and prefrontal cortex in EAS ethnicity. While, \u003cem\u003ePNLIP\u003c/em\u003e is identified only in EAS liver and muscle. Therapeutic potential, as measured by combining coverage, FDR, NES, show significant recovery of GSPs (Figure S10 A (i) and (ii)) within pituitary and ovary showing maximum enrichment.\u003c/p\u003e \u003cp\u003eFurther, we validated our approach by probing enrichment of the DEGS from PCOS patients within the priority matrix (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC (i) and (ii)). Links between PCOS and a few metabolic disorder including obesity and type 2 diabetes (T2D) are well studied\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e and this prompted us to investigate enrichment of disease-specific DEGs (adipose, liver and pancreas) from obesity and T2D patients in the priority gene list of PCOS (Figure S10 B and C (i) and (ii)). While both the up and downregulated DEGs were well enriched in EAS population (NES up to 1.15 in obesity and NES up to 1.74 in T2D), in EUR primarily the upregulated DEGS were enriched.\u003c/p\u003e\n\u003ch3\u003ePathway cross-talk identifies nodal points to perturb multiple pathways:\u003c/h3\u003e\n\u003cp\u003eIn order to complement the individualistic prioritized biological pathways (as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e, 1st Tab-S2, 14th Tab), we set out to identify the interaction between pathways by merging the Reactome pathways on the background interactome (See methods). To this end, we constructed a maximum-scoring subgraph by maximising the number of highly prioritised genes within it along with a few low scoring interconnecting genes. Through this we obtained the crosstalk between pathways (Figure S11 and Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e, 1st Tab-S2, 14th Tab) in different tissues across population. With this exercise, we realized that the high ranking genes in the significance matrix, belonging to different pathways were all well connected with each other (Figure S12 A and B) validating the robustness of the network and significance of our prioritisation strategy. The few low scoring interconnecting genes include \u003cem\u003eCALM1\u003c/em\u003e (Rank 120, Ovary), \u003cem\u003eCDK1\u003c/em\u003e (Rank 159, Pancreas), \u003cem\u003eCYCS\u003c/em\u003e (Rank 106, Cortex) in EAS and \u003cem\u003eCYCS\u003c/em\u003e (Ranking above 160 in Liver, Muscle and Pituitary) in EUR. They are all important in linking the pathways and may represent important nodal points for perturbing multiple pathways together. Connecting these pathways through a minimal spanning tree, where the edges of connections are manifested by the number of shared genes between pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA ii and B ii and Table \u003cspan refid=\"MOESM5\" class=\"InternalRef\"\u003eS5\u003c/span\u003e, 1st Tab), helped us to identify the common important nodal points of perturbation.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe observed several genes like \u003cem\u003eEGFR, KRAS\u003c/em\u003e, and \u003cem\u003eNRAS\u003c/em\u003e, are shared between multiple signalling pathways including signalling by Tyrosine Kinase, Nuclear Receptor in EAS population (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA i, ii and iii). In case of EUR too, genes like \u003cem\u003eRPS27A,TP53\u003c/em\u003e, and \u003cem\u003eUBA52\u003c/em\u003e are shared between majority of the signalling including DNA repair, TAK1-dependent IKK and NF-κB activation, Regulation of TP53 activity (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB i, ii and iii), all of which have been implicated in impaired folliculogenesis and related pathophysiology\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e and could be well utilised to control multiple pathways simultaneously. The crosstalk genes and the pathways they are enriched in for other tissues are presented in Table \u003cspan refid=\"MOESM5\" class=\"InternalRef\"\u003eS5\u003c/span\u003e, 2nd Tab-S5, 7th Tab.\u003c/p\u003e\n\u003ch3\u003eNode removal from crosstalk network reveals resilience and identifies prospects of drug repurposing:\u003c/h3\u003e\n\u003cp\u003eNext, we further explored the contribution of the genes to the robustness and integrity of the crosstalk networks by accessing the tolerance of these network to node removal (Table \u003cspan refid=\"MOESM6\" class=\"InternalRef\"\u003eS6\u003c/span\u003e), where we quantified the fraction of nodes disconnected after removal of a particular node (gene). It stems from an idea that a large number of nodes will be disconnected upon removal of a node critical for the network integrity. We observed that the crosstalk networks were quite tolerant (Table \u003cspan refid=\"MOESM6\" class=\"InternalRef\"\u003eS6\u003c/span\u003e, 1st Tab) to single node removal for both the population (Figure S12 A and B) with the maximum effect achieved through removal of \u003cem\u003eTP53\u003c/em\u003e (~\u0026thinsp;13%) in case of EAS (Figure S12 A). Where as in case of EUR the effect of single node removal was not so prominent (Figure S12 B) with the highest of ~\u0026thinsp;5% effect achieved upon removal of \u003cem\u003eCDK9\u003c/em\u003e. This robustness intrigued us to explore the effect of a combinatorial attack for node removal analysis to find out the smallest possible combination of nodes, ranging between two to four, mimicking a strategy very similar to multi target modulation to mitigate disease. We tried all combination including \u003cem\u003eTP53\u003c/em\u003e and \u003cem\u003eCDK9\u003c/em\u003e respectively for EAS and EUR, which would considerably destabilize the networks. We found that in EAS, \u003cem\u003eTP53\u003c/em\u003e in combination of \u003cem\u003eGNAS\u003c/em\u003e achieved an effect of 23%, whereas, \u003cem\u003eTP53\u003c/em\u003e and \u003cem\u003eGNAS\u003c/em\u003e when additionally combined with \u003cem\u003eYWHAE\u003c/em\u003e, we observed a disconnection of 31% nodes (Figure S12 A). When we further performed the attack analysis with 4 nodes removal, we observed \u003cem\u003eTP53, CTNNB1, RUNX1\u003c/em\u003e and \u003cem\u003eYWHAE\u003c/em\u003e increased effect to 46%. In EUR, two-node removal achieved 7.5%, and with the removal of three-node sets such as \u003cem\u003eCDK9\u003c/em\u003e-\u003cem\u003eCDK1\u003c/em\u003e-\u003cem\u003eELAVL1\u003c/em\u003e or \u003cem\u003eCDK9\u003c/em\u003e-\u003cem\u003eNEIL2-XRCC1\u003c/em\u003e a disconnect of 12.5% was seen. Several four-node sets removal also achieved around 15% disconnect. All these observations were consistent with the idea that attacking multiple targets simultaneously are more effective to perturb resilient modules. Single-node removal in other tissues (Table \u003cspan refid=\"MOESM6\" class=\"InternalRef\"\u003eS6\u003c/span\u003e, 2nd Tab-S6, 7th Tab) identified tissue-specific vulnerabilities, with a highest impact for \u003cem\u003eAPP\u003c/em\u003e removal of 65.7% and 14.6% in adipose and liver respectively in EAS. In EUR however, single-node effects had low to modest effect (Table \u003cspan refid=\"MOESM6\" class=\"InternalRef\"\u003eS6\u003c/span\u003e, 2nd Tab-S6, 7th Tab).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eNode removal analysis further motivated us to evaluate the potential of perturbing the important cross talk genes through existing therapeutics. For the purpose we curated existing therapeutics from ChEMBL. We found considerable support in the potential to perturb a decent number of genes in both the crosstalk networks (EUR and EAS) from clinical evidences of phased and approved drug targets (approved drugs; FDR\u0026thinsp;=\u0026thinsp;1.2\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;4\u003c/sup\u003e and phase 3 drugs for EAS, and approved drugs; FDR\u0026thinsp;=\u0026thinsp;2.1\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e and phase 3 drugs for EUR) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA i and B i respectively). This implies that all of these genes are important targets for other disease indications (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA iii and B iii). We further performed the attack analysis to identify specifically the contribution of these phased drug targets on the network integrity, as a measure of potential of drug repurposing. In alignment with the previous crosstalk found both the networks were robust to single node removal, showing a moderately low effect of attack. However, within the list, \u003cem\u003eTP53\u003c/em\u003e, a phase 3 drug target had the maximum effect of 13% disconnect (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA ii) in EAS. Additionally, in the context of this \u003cem\u003eTP53\u003c/em\u003e removal, combining \u003cem\u003eCALM1\u003c/em\u003e an approved drug target showed a decent 18% disconnect. \u003cem\u003eCALM1\u003c/em\u003e is approved drug target of cardiovascular disease highlighting the link between PCOS and CVD. On the other hand, the node removal analysis did not provide any strong contender that could have a considerable effect on network in case of EUR population (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB ii) with the only exception of \u003cem\u003eCDK9\u003c/em\u003e that have a marginal effect of 5% disconnect. Majority of the targets have similar effect on the network connection displaying a moderate but equal weightage to these drug targets.\u003c/p\u003e\n\u003ch3\u003eNetwork modularity study uncovers tissue-specific and shared cross-tissue functional modules:\u003c/h3\u003e\n\u003cp\u003eThe study of pathway cross talks across tissues intrigued us to capture how different tissues are connected through these pathways across the population. To this end, we merged the pathway crosstalk instances from all the tissues together to obtain functional modules (based on network modularity) from the integrated network (Table \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e, 1st Tab and S7, 2nd Tab) and studied the prominence of different pathways in these modules across tissues. The exploration revealed six (M1-M6) modules in both EAS (Figure S13) and EUR (Figure S14) populations. Without any preconceived assumption and knowledge about functionality, we observed that each of these modules were linked to distinct molecular pathways underlying their association to non-random functional design (Table \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e). We observed both shared and population-specific functional modules across tissues. Among the common pathways enriched by the modules in both the population, signalling by GPCRs, vesicle mediated transport, and translation are the most prominent ones (Table \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e and Figure S13 A and S14 A). Population specific modules are enriched in immune related pathways (M2) and DNA repair (M6) in EUR and various signalling axes (M6) and metabolism (M2) in EAS respectively. When we scored the modules in each of these populations with tissue-specific significance score, an interesting pattern emerged. We observed that the module containing genes related to several hormones (\u003cem\u003eTSHB, FSHB\u003c/em\u003e and \u003cem\u003eCGB\u003c/em\u003e) reportedly released from the pituitary\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e (M4 in both the population), specifically showed high intensity coloration when the modules were coloured by pituitary specific significance score (Figure S13 and S14 B). Moreover, in EAS ovary, highly-rated candidates including \u003cem\u003eFSHR\u003c/em\u003e and \u003cem\u003eLHCGR\u003c/em\u003e from M4 are connected to the majority of signalling genes in M6 via \u003cem\u003eGNAS\u003c/em\u003e. A couple of these genes are also prominent members of Hypothalamus-pituitary axis\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Some other important genes specifically highlighted in pituitary only includes \u003cem\u003eDRD1, DRD5, HTR6\u003c/em\u003e and \u003cem\u003ePOMC\u003c/em\u003e genes, thus making the M4 truly exclusive for pituitary tissue. Signalling related to Vesicular Transport is enriched in module prioritized in muscle (Figure S13 and S14 B, Table \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e), whereas the translation and RNA related metabolism are highlighted in the cortex. Interestingly, immune centric module M2 was specifically highlighted in liver and in alignment to this, a few important innate immunity genes containing module M6 (\u003cem\u003eIRF1, IFNA21, IFNA10, JAK, STAT3\u003c/em\u003e etc.) was collectively prioritized in ovary, pancreas and adipose tissue in the EUR population. Collectively, these finding adds an additional level of validation for the relevance and significance of the scoring strategy and the pathway crosstalk.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eSelf-organizing map identifies physiological similarity within tissues and defines therapeutic space:\u003c/h2\u003e \u003cp\u003eTo understand more about the shared and unique cross talk genes between different tissues of both the population, we did a pairwise correlation analysis between tissue-specific significance scores from both the population (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA). We found weak or negative correlations between significance scores of tissues, except for a modest positive correlation for cortex and pituitary (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA). This underlines again a prominent distinction between the prioritized genes between the populations. To further understand this distinction and to capture the similarity between these tissues, we proceeded to find the genes that are shared between the tissues within a population. This population centric summary of the shared and unique genes participating in the crosstalk is important to define a common therapeutic space for multiple tissues involved in the disease. To this end, we used an unsupervised self-organising map approach to cluster the crosstalk genes in to a self-organized map (SOM) thereby depicting similar pattern of significance scoring in a cluster (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB (i) and C (i)). Tissue specific maps thus created placed tissues with similar profiles close to one another (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB (i) and 6C(i)). This revealed that there are both similarity and variability with respect to the significance scoring patterns of the cross-talk genes between tissues. Pituitary and adipose displayed high similarity in EAS in contrast to EUR population. Whereas, ovary and muscle had a higher similarity in clustering patterns in EUR when compared EAS. However the pancreas had a moderately similar distribution of the scoring pattern. These observation intrigued us to explore further to find out clusters in different tissues having similar scoring pattern and to create a consolidated supra-hexagonal cluster repressing all the tissues for a population. In the process, we identified 4 and 5 self-organising clusters in both EAS and EUR respectively (Figure S15) containing genes with similar pattern of significance scoring. The highly scored clusters emerged to be C3 and C2in EAS and EUR population respectively. These clusters highly scored genes in all the tissues analysed as displayed in the ridge plots with the score density mapping (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB (iv) and C (iv)). Interesting, for the EAS and EUR population, the same cluster happens to also contain the genes which are highly druggable as measured by the number of druggable pockets (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB (iii \u0026amp; v) and C (iii \u0026amp; v)).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo get a biological overview of the clusters, we examined the pathways of these clusters. In this regard we performed pathway enrichment with the genes present in the druggable as well as high scored clusters (Figure S16 \u0026amp; Table \u003cspan refid=\"MOESM8\" class=\"InternalRef\"\u003eS8\u003c/span\u003e). In both the population pathways were very similar as reported in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. Specifically, pathways related to signalling by receptor tyrosine kinase, nuclear receptors including Estrogen Receptor and signalling regulating membrane trafficking were prominently enriched in the high druggability and highly scored cluster \u0026lsquo;C3\u0026rsquo; in the EAS (Figure S16A \u0026amp; Table \u003cspan refid=\"MOESM8\" class=\"InternalRef\"\u003eS8\u003c/span\u003e, 1st Tab). While, the Cluster C2 in EUR representing both druggable and well scored cluster, enriched in signalling related to tyrosine kinase, cell cycle and DNA repair (Figure S16B \u0026amp; Table \u003cspan refid=\"MOESM8\" class=\"InternalRef\"\u003eS8\u003c/span\u003e, 2nd Tab). Enrichment of pathways related to several immune related function was also in alignment with our previous observations (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003ePrioritized druggable clusters reveal drug repurposing prospects across populations:\u003c/h3\u003e\n\u003cp\u003eTo explore the prospects of drug repurposing and also to assess the overall therapeutic landscape across tissues and population, we represented the cluster with high druggablility and significance scores in both the population as heat maps depicting their significance score (Figs.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e and \u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). The illustration also includes how these genes are affiliated to different reactome pathways. We observed that within the highly scored cluster C3 in EAS, the H29 displayed higher distribution of significance score for most of the genes in that hexagon across the tissues (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). The genes in this hexagonal unit predominantly belongs to signalling by RTKs and Membrane trafficking. Few important genes in this hexagon include \u003cem\u003eERBB3, RAB5B, CDK2\u003c/em\u003e etc. which are involved in pathways related to PCOS pathologies\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. We found approved drugs targeting \u003cem\u003eERBB3\u003c/em\u003e and \u003cem\u003eRPS26\u003c/em\u003e, which are prescribed for various cancers and muscular dystrophy respectively. The unit, H27 containing \u003cem\u003eFSHR\u003c/em\u003e, a well-known candidate significantly involved in the biology pertaining to follicular development, also appeared as one of the promising targets. It is already in clinical investigation for PCOS being targeted by drugs Follitropin (phase 2) and Menotropin (phase3). The hexagonal unit H31 emerged as a critical hub with genes that are enriched in multiple pathways involving signalling by RTKs and nuclear receptors (ESR mediated signalling). This cluster includes approved drug targets and targets with high number of available druggable PDB structures including \u003cem\u003eIGF1R, TP53, GRB2\u003c/em\u003e and \u003cem\u003eCTNNB1\u003c/em\u003e. In the EUR population on the other hand (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e), the highly scored cum druggable cluster (cluster 2) is strongly associated with immune related pathways. The cluster H20, displayed high significance score across multiple tissues with approved drugs targets involved in of cell cycle, transcription and immune response pathways. The H37 containing highly scored \u003cem\u003eIRF1\u003c/em\u003e in majority of the metabolic tissues, is reported to be an important player in innate immune response. Similarly, in H8 presence of \u003cem\u003eSTAT1, NF-\u003c/em\u003eκ\u003cem\u003eB1\u003c/em\u003e and \u003cem\u003eRELA\u003c/em\u003e are strongly associated with pathways related to immune response. All these observation signifies the potential of these highly scored clusters for further exploration with respect to drug discovery.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003ePCOS affects around 10% to 13% women worldwide but still a lack of in depth understanding about the aetiology and the underlying mechanism of the disease manifestation, have restricted the development of an effective therapy of PCOS. However, some reports suggests that genetic, epigenetic and environmental factors contribute collectively towards the disease manifestation. Till date an extensive study towards the understanding of the interplay between the genetic and epigenetic regulation that might lead to the disease development and the pathophysiology, is lacking. Therefore, in this study we integrated PCOS-associated genetic variants, tissue-specific regulatory genomics and protein interaction network information to better understand the disease biology. Specifically, the inclusion of ChromHMM, genomic conformation and genome-wide QTL summary statistics (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) to our study enabled us to leverage the regulatory annotations of the genomic variants. As majority of these data are tissue and population specific, it gave us a way to understand how disease biology is different due to different ethnic background and different disease relevant tissues. Scoring lead us to prioritize different genetic targets in EAS and EUR population. Inclusion of protein interaction information, resulted in prioritizing an additional list of secondary genetic targets, which happen to be closely associated with the primary genetic targets having genetic evidence, as in \u0026ldquo;guilt by association\u0026rdquo;. Indeed, for example in ovary, we noticed various important genes with respect to PCOS biology like \u003cem\u003eIGF1R, INS, EGFR\u003c/em\u003e, \u003cem\u003eESR1, NF-\u003c/em\u003eκ\u003cem\u003eB1, AMH\u003c/em\u003e and \u003cem\u003eAR\u003c/em\u003e were picked up because of network connectivity. Eventually, many important signaling pathways were predominantly enriched due to these peripheral genes. Interestingly, differential scoring in different population revealed a preferential prioritization of immune related pathways in EUR. Whereas, metabolic and hormonal dysfunction are mostly predominant in the priority list of the EAS population with \u003cem\u003eINSR, IGF1R, LHCGR, FSHR\u003c/em\u003e scoring high when compared to EUR prioritization. It is worthwhile to mention that PCOS has been linked to a systemic low grade inflammation\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. To this end, we investigated the pattern of immune, metabolic and hormonal dysregulation in PCOS patients from a few gene expression dataset. In alignment to our prioritization we also observed a statistically significant preferential enrichment of immune related pathways in EUR and metabolic and hormonal pathways in EAS. This observation, on one hand underlined the difference in PCOS manifestation in different population, thereby indicating why we might need personalized strategies for PCOS management for different ethnicities. This observations also validated our scoring strategy. In fact, our strategy outperforms few other equivalent strategies of genetic target prioritization. Together, our findings highlight several key differences in etiology between populations. Among the important signaling in EAS population, signaling by receptor tyrosine kinases and insulin receptor are also reported elsewhere in association to PCOS. Additionally, prioritized genes in EAS ovary showed pronounced enrichment of multiple signal transduction pathways (Figure \u003cspan refid=\"MOESM5\" class=\"InternalRef\"\u003eS5\u003c/span\u003e) including some relevant receptor tyrosine kinases (RTKs) (insulin-like growth factor 1 receptor and insulin receptor). All these suggest that EAS PCOS might be characterized by hyper activation of insulin/IGF-1-mediated metabolic and proliferative signaling cascades. The metabolic dysregulation extends beyond the ovary as evidenced by prioritization of insulin receptor signaling cascade in all the metabolic tissue and leptin signaling in liver\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. The prominence of these pathways aligns with clinical observations that EAS women with PCOS exhibit more severe metabolic dysfunction and increased risk of metabolic complications compared to their EUR counterparts\u003csup\u003e\u003cspan additionalcitationids=\"CR37\" citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. The enrichment of hormone ligand-binding receptors across adipose, liver, pituitary and ovary tissues, peptide hormone biosynthesis (adipose, liver and pituitary) and androgen biosynthesis (adipose, pituitary and ovary) suggest systemic hormonal dysfunction in EAS population. Similarly, the prioritization of the TLR4-NF-κB-NLRP3 signaling axis in EUR population, further offers a specific molecular hypothesis for this inflammatory phenotype unique to the population. This inflammatory signature may relate to the observation that EUR ancestry PCOS patients often present with systemic inflammation\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e,\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. The systemic inflammatory profile is further evidenced by shared enrichment of IL21 and IL27 signaling across adipose, pancreas, and muscle, with IFNα and IFNβ signaling predominantly enriched in adipose tissue. The shared enrichment of ERBB2 signaling and PTK6 across both populations in most of the tissues suggests that certain core mechanisms is conserved regardless of ethnic background. We obtained an additional validation when using our priority list, we could retrieve a few potential PCOS drug targets that are being actively investigated in several stages of clinical trials. Enrichment of differentially expressed genes of obese and T2D patients in the prioritized gene list on the other hand, underscores the biological relevance of the PCOS associated co-morbidities. Our finding also establishes link though important genetic targets through which PCOS patients might get susceptible for many other metabolic disorders and cancer. Specifically, the role of several pathways involving protein tyrosine kinase, signaling by insulin receptors, inflammatory signaling etc. are implicated in both obesity and T2D\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. All of these pathways were seen to be enriched in the context of PCOS also (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e, 1st Tab-S2, 14th Tab). This observation warrants a further study towards the pleotropic effect of genetic variants which might lead to several associated co-morbidities including cardiovascular disease, hypertension, cancer, infertility and psychological disorders. In an attempt to explore more into the underlying mechanism of how different signaling pathways interact with each other, we identified the common genes through which these pathways cross talk. These common genes then become the hub to control multiple pathways together by targeting them by drugs. Identification of these target genes (nodes) and assessing the importance of these nodes through node removal analysis, gave us an idea of how some of these targets can be used for drug repurposing and to what extent these repurposing could be meaningful. From this exercise, we also realized that most of these genes are highly connected with each other and thus it is not so easy to perturb the disease network by targeting a single node. In some cases, combinatorial removal analysis was able to achieve a better effect, underlining the robustness of these networks, which eventually supports a polygenic model of disease manifestation\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e, where subtle expression changes of numerous target genes collectively shape disease manifestation. These findings also reinforce the notion that crosstalk between disease networks in complex traits are characterized by high interconnectivity and redundancy, making them intrinsically robust to perturbation. Pathway crosstalk also led us to explore cross tissue cross-talk which is an important step in understanding how the disease-relevant tissues collaborate with each other to manifest PCOS pathophysiology. We explored the modularity of the resulting integrated cross-tissue network and observed a few interesting modules enriched in specific tissues. Modules scoring high in cortex in both the population is enriched predominantly in cellular response to starvation and translation pathways. As it is reported that PCOS is often associated with reduced AMPK activity and parts of brain cortex and hypothalamus is responsible for starvation and hunger homeostasis, a connection to this pathway can be conceived as connection to metabolic imbalance and disruption of ovarian function\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. The other important observation include the module enriched in pituitary specific TSHB and FSHB prioritization and high scoring of immune related genes in metabolic tissues in EUR, highlighting the involvement of these tissues in chronic inflammation\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Finally, the supra-hexagonal clustering provided an elegant, data-driven method to identify promising therapeutic targets by integrating tissue-wide similarity in scoring pattern and thus identifying potential therapeutic space and repurposing opportunities. For example, we could retrieve a few candidates like Follitropin and Menotropin targeting FSHR and Flutamide targeting AR (Figs.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e and \u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e) which are under active clinical investigation. In a nutshell, our study focused on connecting the genetic variants to regulatory implications and thereby linking them to the pathophysiology of PCOS. Further, with differential prioritization of genetic targets in population and tissues we understood the subtle similarities and differences of the disease manifestation and thus opening up avenues to understand and treat the disease at a personalized scale. Thus our study redefines the ethnic heterogeneity in PCOS from being a mere confounding variable into a biologically informative dimension. Exploring how the disease related signaling propagates through various signaling axes and integrating tissue-wise patterns led us to define common therapeutic space, which needs to be explored further to develop new generation drugs for PCOS.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eThe pipeline:\u003c/h2\u003e \u003cp\u003eThe overview of the entire flow has been illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Population-specific GWAS summary statistics relevant to PCOS were systematically curated from publicly available database (see \u0026ldquo;\u003cb\u003eGWAS summary-level data collection and processing\u003c/b\u003e\u0026rdquo;). To identify genes potentially influenced by PCOS-associated single nucleotide polymorphisms (SNPs), we employed three complementary approaches: (i) rGenes: Genes identified due to physical proximity to SNPs integrated with epigenetic regulatory information (ii) qGenes: Genes identified due to associations of the SNPs with a quantitative genomic trait (QTL) (iii) cGenes: Genes identified due to SNPs involved in chromatin interactions (see \u0026lsquo;\u003cb\u003eIdentification of core genes under genomic influence\u003c/b\u003e\u0026rsquo;). Together, these approaches identified a set of \u0026ldquo;core genes\u0026rdquo;, each supported by direct genomic evidence linking them to PCOS-associated SNPs. These genes were further annotated based on their relevance to PCOS, considering whether they had previously been implicated in the syndrome or associated with PCOS-related phenotypes (see \u0026lsquo;\u003cb\u003eAnnotation of core genes with functional evidences\u003c/b\u003e\u0026rsquo;). Next, we expanded the scope of gene identification by considering genes potentially influenced by network interactions. Using a curated Protein-Protein Interaction (PPI) network, we identified \u0026ldquo;peripheral genes\u0026rdquo;, subsequently both the core and the peripheral genes and were quantified based on their connectivity to each other using Random Walk algorithm (see \u0026lsquo;\u003cb\u003eIdentification of peripheral genes with network evidence\u003c/b\u003e\u0026rsquo;). The core and peripheral genes, along with their genomic and network evidences were combined into a gene-predictor matrix, where each gene was assigned a composite score ranging from 0 to 5 to reflect its overall clinical importance with respect to PCOS (see \u0026lsquo;\u003cb\u003eGene-predictor matrix to gene prioritization\u003c/b\u003e\u0026rsquo;). This comprehensive approach was applied specifically to PCOS-relevant tissues including the Ovary, Adipose, Liver, Muscle, Pancreas, Brain and Pituitary, ensuring that the prioritization process captured tissue-specific regulatory differences and population-level variation relevant to PCOS pathophysiology. To further investigate the functional roles of prioritized genes, we integrated pathway-level information into the PPI network, enabling the identification of genes that mediate interactions between distinct biological pathways (see \u0026lsquo;\u003cb\u003epathway crosstalk\u003c/b\u003e\u0026rsquo;). Next, we performed cluster analysis on the crosstalk-mediating genes across tissues to identify gene clusters having similarity in scoring pattern across tissues that may represent shared pathophysiology and define potential therapeutic space for PCOS (see \u0026lsquo;\u003cb\u003eCluster analysis\u003c/b\u003e\u0026rsquo;).\u003c/p\u003e \u003cp\u003eThe effectiveness of the prioritization framework was assessed using two performance evaluation strategies. We benchmarked the framework\u0026rsquo;s ability to distinguish GSPs from GSNs in a tissue-specific context. (See \u0026lsquo;\u003cb\u003eBenchmarking the scoring strategy\u003c/b\u003e\u0026rsquo;). We examined the prioritization framework\u0026rsquo;s potential to retrieve the known therapeutics of PCOS (see \u003cb\u003e\u0026lsquo;Target set enrichment analysis (Validation 1)\u003c/b\u003e and \u003cb\u003e\u0026lsquo;Genetics-to-Current-Therapeutics (G2CT) potential\u0026rsquo;\u003c/b\u003e). A further validation was performed by gene set enrichment analysis of differentially expressed genes from PCOS patients. Detailed Materials and Methods are\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eS.R. collected the data, carried out the analyses according to the adopted pipeline. \u0026nbsp;D.D. framed the research question, analyzed the data, supervised the work, and wrote the manuscript. All authors approved the final version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThere was no funding associated with this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDisclosure statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo potential conflict of interest was reported by the author(s).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData sharing is not applicable to this study, as no new data were created. However, the authors confirm that the data supporting the findings of this study are available within the article, its supplementary materials, and referenced publications from which the data were extracted.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eStener-Victorin, E. \u003cem\u003eet al.\u003c/em\u003e Polycystic ovary syndrome. \u003cem\u003eNat. Rev. Dis. Prim. 2024 101\u003c/em\u003e 10, 27- (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAkre, S., Sharma, K., Chakole, S. \u0026amp; Wanjari, M. B. Recent Advances in the Management of Polycystic Ovary Syndrome: A Review Article. \u003cem\u003eCureus\u003c/em\u003e 14, e27689 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVink, J. M., Sadrzadeh, S., Lambalk, C. B. \u0026amp; Boomsma, D. I. Heritability of polycystic ovary syndrome in a Dutch twin-family study. \u003cem\u003eJ. Clin. Endocrinol. Metab.\u003c/em\u003e 91, 2100\u0026ndash;2104 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLegro, R. S., Driscoll, D., Strauss, J. F., Fox, J. \u0026amp; Dunaif, A. Evidence for a genetic basis for hyperandrogenemia in polycystic ovary syndrome. \u003cem\u003eProc. Natl. Acad. Sci. U. S. A.\u003c/em\u003e 95, 14956 (1998).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDay, F. et al. Large-scale genome-wide meta-analysis of polycystic ovary syndrome suggests shared genetic architecture for different diagnosis criteria. PLoS Genet. (2018) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pgen.1007813\u003c/span\u003e\u003cspan address=\"10.1371/journal.pgen.1007813\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCook, D. \u003cem\u003eet al.\u003c/em\u003e Lessons learned from the fate of AstraZeneca\u0026rsquo;s drug pipeline: a five-dimensional framework. \u003cem\u003eNat. Rev. Drug Discov.\u003c/em\u003e 13, 419\u0026ndash;431 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNelson, M. R. \u003cem\u003eet al.\u003c/em\u003e The support of human genetic evidence for approved drug indications. \u003cem\u003eNat. Genet.\u003c/em\u003e 47, 856\u0026ndash;860 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWaring, M. J. \u003cem\u003eet al.\u003c/em\u003e An analysis of the attrition of drug candidates from four major pharmaceutical companies. \u003cem\u003eNat. Rev. Drug Discov.\u003c/em\u003e 14, 475\u0026ndash;486 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUffelmann, E. \u003cem\u003eet al.\u003c/em\u003e Genome-wide association studies. \u003cem\u003eNat. Rev. Methods Prim. 2021 11\u003c/em\u003e 1, 59- (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJi, S. Y. \u003cem\u003eet al.\u003c/em\u003e The polycystic ovary syndrome-associated gene Yap1 is regulated by gonadotropins and sex steroid hormones in hyperandrogenism-induced oligo-ovulation in mouse. \u003cem\u003eMol. Hum. Reprod.\u003c/em\u003e 23, 698\u0026ndash;707 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCensin, J. C., Bovijn, J., Holmes, M. V. \u0026amp; Lindgren, C. M. Colocalization analysis of polycystic ovary syndrome to identify potential disease-mediating genes and proteins. \u003cem\u003eEur. J. Hum. Genet.\u003c/em\u003e 29, 1446 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, J. H. \u003cem\u003eet al.\u003c/em\u003e Role of EGFR expressed on the granulosa cells in the pathogenesis of polycystic ovarian syndrome. \u003cem\u003eFront. Endocrinol. (Lausanne).\u003c/em\u003e 13, 971564 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLaban, M. A., Shemais, T. A., Elhelw, E. M. \u0026amp; El-Hassanin, A. R. Role Of Insulin Like Growth Factor in Polycystic Ovary Syndrome. \u003cem\u003eEgypt. J. Hosp. Med.\u003c/em\u003e 72, 4005 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTan, W. \u003cem\u003eet al.\u003c/em\u003e Insights on the NF-κB system in polycystic ovary syndrome, attractive therapeutic targets. \u003cem\u003eMol. Cell. Biochem.\u003c/em\u003e 479, 467\u0026ndash;486 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRudnicka, E. \u003cem\u003eet al.\u003c/em\u003e Chronic Low Grade Inflammation in Pathogenesis of PCOS. \u003cem\u003eInt. J. Mol. Sci.\u003c/em\u003e 22, 3789 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBremer, A. A. \u0026amp; Miller, W. L. The serine phosphorylation hypothesis of polycystic ovary syndrome: a unifying mechanism for hyperandrogenemia and insulin resistance. \u003cem\u003eFertil. Steril.\u003c/em\u003e 89, 1039\u0026ndash;1048 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClark, K. L. \u003cem\u003eet al.\u003c/em\u003e Hippo Signaling in the Ovary: Emerging Roles in Development, Fertility, and Disease. \u003cem\u003eEndocr. Rev.\u003c/em\u003e 43, 1074\u0026ndash;1096 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHaouzi, D. \u003cem\u003eet al.\u003c/em\u003e Altered gene expression profile in cumulus cells of mature MII oocytes from patients with polycystic ovary syndrome. \u003cem\u003eHum. Reprod.\u003c/em\u003e 27, 3523\u0026ndash;3530 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZiemak, H. \u003cem\u003eet al.\u003c/em\u003e Angiogenesis in the ovary \u0026ndash; Proangiogenic factors and their clinical applications. \u003cem\u003eBiomed. Pharmacother.\u003c/em\u003e 191, 118416 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, J., Yin, T. \u0026amp; Liu, S. Dysregulation of immune response in PCOS organ system. \u003cem\u003eFront. Immunol.\u003c/em\u003e 14, 1169232 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo, Q. \u003cem\u003eet al.\u003c/em\u003e NF-κB in biology and targeted therapy: new insights and translational implications. \u003cem\u003eSignal Transduct. Target. Ther.\u003c/em\u003e 9, 1\u0026ndash;37 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu, Q. \u0026amp; Kanneganti, T.-D. Distinct regulatory mechanisms control proinflammatory cytokines IL-18 and IL-1β. \u003cem\u003eJ. Immunol.\u003c/em\u003e 198, 4210 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFernandes, F. P., Leal, V. N. C., Souza de Lima, D., Reis, E. C. \u0026amp; Pontillo, A. Inflammasome genetics and complex diseases: a comprehensive review. \u003cem\u003eEur. J. Hum. Genet.\u003c/em\u003e 28, 1307\u0026ndash;1321 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu, Y. \u0026amp; Qiao, J. Association of Insulin Resistance and Elevated Androgen Levels with Polycystic Ovarian Syndrome (PCOS): A Review of Literature. \u003cem\u003eJ. Healthc. Eng.\u003c/em\u003e 2022, 9240569 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown, J. \u0026amp; Farquhar, C. Clomiphene and other antioestrogens for ovulation induction in polycystic ovarian syndrome. \u003cem\u003eCochrane database Syst. Rev.\u003c/em\u003e 12, (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAustreg\u0026eacute;silo de Athayde De Hollanda Morais, B. \u003cem\u003eet al.\u003c/em\u003e The efficacy and safety of GLP-1 agonists in PCOS women living with obesity in promoting weight loss and hormonal regulation: A meta-analysis of randomized controlled trials. \u003cem\u003eJ. Diabetes Complications\u003c/em\u003e 38, 108834 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKakoly, N. S. \u003cem\u003eet al.\u003c/em\u003e Ethnicity, obesity and the prevalence of impaired glucose tolerance and type 2 diabetes in PCOS: a systematic review and meta-regression. \u003cem\u003eHum. Reprod. Update\u003c/em\u003e 24, 455\u0026ndash;467 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXue, H. \u003cem\u003eet al.\u003c/em\u003e The mechanism of NF-κB-TERT feedback regulation of granulosa cell apoptosis in PCOS rats. \u003cem\u003ePLoS One\u003c/em\u003e 19, e0312115 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcCartney, C. R., Campbell, R. E., Marshall, J. C. \u0026amp; Moenter, S. M. The role of gonadotropin-releasing hormone neurons in polycystic ovary syndrome. \u003cem\u003eJ. Neuroendocrinol.\u003c/em\u003e 34, e13093 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiao, B., Qiao, J. \u0026amp; Pang, Y. Central Regulation of PCOS: Abnormal Neuronal-Reproductive-Metabolic Circuits in PCOS Pathophysiology. \u003cem\u003eFront. Endocrinol. (Lausanne)\u003c/em\u003e. 12, 667422 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePadder, K. A., Yousuf, M. A., Jahan, N., Yousuf, S. D. \u0026amp; Ganie, M. A. Interplay between elevated RAB5B gene expression and insulin resistance among women with PCOS-insights from a case-control study. \u003cem\u003eEndocrine\u003c/em\u003e 88, 323\u0026ndash;329 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlan Harris, R. \u003cem\u003eet al.\u003c/em\u003e Loci on chromosome 12q13.2 encompassing ERBB3, PA2G4 and RAB5B are associated with polycystic ovary syndrome. \u003cem\u003eGene\u003c/em\u003e 852, 147062 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEscobar-Morreale, H. F., Luque-Ram\u0026iacute;rez, M. \u0026amp; Gonz\u0026aacute;lez, F. Circulating inflammatory markers in polycystic ovary syndrome: A systematic review and metaanalysis. \u003cem\u003eFertil. Steril.\u003c/em\u003e 95, (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAboeldalyl, S. \u003cem\u003eet al.\u003c/em\u003e The Role of Chronic Inflammation in Polycystic Ovarian Syndrome\u0026mdash;A Systematic Review and Meta-Analysis. \u003cem\u003eInt. J. Mol. Sci.\u003c/em\u003e 22, 2734 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeng, Y. \u003cem\u003eet al.\u003c/em\u003e Elevated Serum Leptin Levels as a Predictive Marker for Polycystic Ovary Syndrome. \u003cem\u003eFront. Endocrinol. (Lausanne).\u003c/em\u003e 13, 845165 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo, M. \u003cem\u003eet al.\u003c/em\u003e Comparison of the phenotype of Chinese versus Dutch Caucasian women presenting with polycystic ovary syndrome and oligo/amenorrhoea. \u003cem\u003eHum. Reprod.\u003c/em\u003e 27, 1481\u0026ndash;1488 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLo, J. C. \u003cem\u003eet al.\u003c/em\u003e Epidemiology and Adverse Cardiovascular Risk Profile of Diagnosed Polycystic Ovary Syndrome. \u003cem\u003eJ. Clin. Endocrinol. Metab.\u003c/em\u003e 91, 1357\u0026ndash;1363 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChahal, N. \u003cem\u003eet al.\u003c/em\u003e Comparison of metabolic syndrome elements in White and Asian women with polycystic ovary syndrome: results of a regional, American cross-sectional study. \u003cem\u003eF\u0026amp;S Reports\u003c/em\u003e 1, 305\u0026ndash;313 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTian, D., Chen, J. \u0026amp; Liu, L. Causal relationship between inflammatory cytokines and polycystic ovary syndrome: a bidirectional mendelian randomization study. \u003cem\u003eJ. Ovarian Res.\u003c/em\u003e 17, 1\u0026ndash;7 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRudnicka, E. \u003cem\u003eet al.\u003c/em\u003e Inflammatory Markers in Women with Polycystic Ovary Syndrome. \u003cem\u003eBiomed Res. Int.\u003c/em\u003e 2020, 4092470 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRohm, T. V., Meier, D. T., Olefsky, J. M. \u0026amp; Donath, M. Y. Inflammation in obesity, diabetes, and related disorders. \u003cem\u003eImmunity\u003c/em\u003e 55, 31\u0026ndash;55 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHotamisligil, G. S., Budavari, A., Murray, D. \u0026amp; Spiegelman, B. M. Reduced tyrosine kinase activity of the insulin receptor in obesity-diabetes. Central role of tumor necrosis factor-alpha. \u003cem\u003eJ. Clin. Invest.\u003c/em\u003e 94, 1543\u0026ndash;1549 (1994).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoyle, E. A., Li, Y. I. \u0026amp; Pritchard, J. K. An expanded view of complex traits: from polygenic to omnigenic. \u003cem\u003eCell\u003c/em\u003e 169, 1177 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCao, Z. \u003cem\u003eet al.\u003c/em\u003e Glycolytic Dysfunction in Granulosa Cells and Its Contribution to Metabolic Dysfunction in Polycystic Ovary Syndrome. \u003cem\u003eDrug Des. Devel. Ther.\u003c/em\u003e 19, 5255 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRudnicka, E. \u003cem\u003eet al.\u003c/em\u003e Inflammatory Markers in Women with Polycystic Ovary Syndrome. \u003cem\u003eBiomed Res. Int.\u003c/em\u003e 2020, 4092470 (2020).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"npj-systems-biology-and-applications","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"npjsba","sideBox":"Learn more about [npj Systems Biology and Applications](http://www.nature.com/npjsba/)","snPcode":"41540","submissionUrl":"https://submission.springernature.com/new-submission/41540/3","title":"npj Systems Biology and Applications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Polycystic ovary syndrome, variant to gene mapping, epigenetic regulation, clinical-proof-of-concept, inflammation, differential molecular pathology, therapeutic landscape, drug repositioning","lastPublishedDoi":"10.21203/rs.3.rs-8610143/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8610143/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePolycystic ovary syndrome (PCOS) is a highly prevalent and heterogeneous endocrine disorder affecting women of reproductive age, with substantial reproductive, metabolic, and long-term health consequences. While genome-wide association studies (GWAS) have identified multiple PCOS-associated loci across diverse populations, the functional interpretation of these predominantly non-coding variants and their translation into clinically actionable targets remain unresolved. Here, we present an integrative population-aware framework that systematically combines regulatory functional genomics, long-range chromatin interactions, genome-wide quantitative trait loci, and protein\u0026ndash;protein interaction networks to prioritize effector genes underlying PCOS susceptibility. Applying this framework to East Asian and European populations, we demonstrate robust performance relative to existing approaches and uncover both shared and population-specific functions. Notably, our analyses reveal a predominant enrichment of metabolic dysregulation\u0026ndash;associated pathways in East Asian PCOS, whereas European PCOS exhibits a stronger inflammatory and immune-related signature. These population-specific molecular phenotypes were further supported by transcriptomic data from PCOS patient samples. Importantly, integration of genetic evidence with network-based approach enabled the identification of druggable targets lacking direct genetic cues. Collectively, our study provides mechanistic insight into the ethnic heterogeneity of PCOS and establishes a scalable strategy for genetically informed, population-specific therapeutic prioritization, advancing precision medicine approaches for women\u0026rsquo;s health.\u003c/p\u003e","manuscriptTitle":"Evidence-based genetic variants to gene mapping and prioritization uncovers distinct molecular pathophysiology and therapeutic landscape in polycystic ovary syndrome patients of different ethnicities.","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-22 07:06:02","doi":"10.21203/rs.3.rs-8610143/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-04T06:07:11+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-03T20:45:57+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-01T15:12:11+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"204106757474156193140133071161569891525","date":"2026-03-27T16:29:04+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"158671351833946120247659917534595001132","date":"2026-03-27T05:13:31+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"102098076628577373719878901020133856786","date":"2026-01-21T19:23:44+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"216658212070464316046450828763165204321","date":"2026-01-19T14:40:05+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-01-19T11:39:46+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-19T11:37:41+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-01-19T11:03:14+00:00","index":"","fulltext":""},{"type":"submitted","content":"npj Systems Biology and Applications","date":"2026-01-15T11:40:16+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"npj-systems-biology-and-applications","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"npjsba","sideBox":"Learn more about [npj Systems Biology and Applications](http://www.nature.com/npjsba/)","snPcode":"41540","submissionUrl":"https://submission.springernature.com/new-submission/41540/3","title":"npj Systems Biology and Applications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"2e6ba872-5b29-4de6-9702-67cea6469ac1","owner":[],"postedDate":"January 22nd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":61400658,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":61400659,"name":"Health sciences/Diseases"},{"id":61400660,"name":"Biological sciences/Genetics"}],"tags":[],"updatedAt":"2026-05-13T07:38:39+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-22 07:06:02","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8610143","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8610143","identity":"rs-8610143","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}