{"paper_id":"199fa1b9-4281-4270-ab0b-5b705010269b","body_text":"A multi-level investigation of the genetic relationship \nbetween endometriosis and ovarian cancer histotypes \n \nSally Mortlock1, Rosario I. Corona2, Pik Fang Kho3,4, Paul Pharoah5,6, Ji-Heui Seo7, Matthew \nL. Freedman7,8, Simon A. Gayther9, Matthew T. Siedhoff10, Peter A.W. Rogers11, Ronald \nLeuchter12, Christine S. Walsh13, Ilana Cass12, Beth Y. Karlan14, B.J. Rimel12, Ovarian \nCancer Association Consortium, International Endometriosis Genetics Consortium, Grant W. \nMontgomery1, Kate Lawrenson**2,9,12, Siddhartha P. Kar**15,16. \n \n1The Institute for Molecular Bioscience, The University of Queensland, Brisbane, QLD 4072, \nAustralia.  \n2Women's Cancer Research Program at Samuel Oschin Comprehensive Cancer Institute, \nCedars-Sinai Medical Center, Los Angeles, CA, USA.  \n3Department of Genetics and Computational Biology, QIMR Berghofer Medical Research \nInstitute, Brisbane, Queensland, Australia. \n4School of Biomedical Science, Faculty of Health, Queensland University of Technology, \nBrisbane, Queensland, Australia. \n5Centre for Cancer Genetics Epidemiology, Department of Public Health and Primary Care, \nUniversity of Cambridge, Cambridge, CB1 8RN, UK. \n6Centre for Cancer Genetics Epidemiology, Department of Oncology, University of \nCambridge, Cambridge, CB1 8RN, UK. \n7Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, USA. \n8Center for Functional Cancer Epigenetics, Dana-Farber Cancer Institute, Boston, MA, USA. \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \nNOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.\n\n9Center for Bioinformatics and Functional Genomics, Samuel Oschin Comprehensive Cancer \nInstitute, Cedars-Sinai Medical Center, Los Angeles, CA, USA. \n10Division of Minimally Invasive Gynecologic Surgery, Department of Obstetrics and \nGynecology, Cedars-Sinai Medical Center, Los Angeles, CA, USA. \n11University of Melbourne Department of Obstetrics and Gynaecology, and Gynaecology \nResearch Centre, Royal Women’s Hospital, Parkville VIC 3052, Australia. \n12Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Cedars-\nSinai Medical Center, Los Angeles, CA, USA. \n13Department of Obstetrics and Gynecology, University of Colorado, Aurora, Colorado, USA \n14Department of Obstetrics and Gynecology, David Geffen School of Medicine, University of \nCalifornia, Los Angeles, CA, USA. \n15Medical Research Council Integrative Epidemiology Unit, University of Bristol, Bristol, \nBS8 2BN, UK. \n16Population Health Sciences, Bristol Medical School, University of Bristol, Bristol, BS8 \n2BN, UK. \n \n**These authors jointly supervised this study and are co-last authors. \n \nAbstract \nEndometriosis is associated with increased risk of epithelial ovarian cancers (EOCs). Data \nfrom large endometriosis and EOC genome-wide association meta-analyses were used to \nestimate the genetic correlation and evaluate the causal relationship between genetic liability \nto endometriosis and major EOC histotypes, and to identify shared susceptibility loci. We \nestimated a significant genetic correlation (rg) between endometriosis and clear cell (rg=0.71), \nendometrioid (rg=0.48) and high-grade serous (rg=0.19) ovarian cancer, all supported by \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nMendelian randomization analyses. A bivariate meta-analysis identified 28 loci associated \nwith endometriosis and EOC, including two novel risk loci, and 19 with evidence for a single \nunderlying causal variant. Differences in the shared risk suggest different underlying \npathways may contribute to the relationship between endometriosis and the different \nhistotypes. Functional annotation using transcriptomic and epigenomic profiles of relevant \ntissues/cells highlighted several target genes. This comprehensive analysis reveals profound \ngenetic overlap between endometriosis and EOC histotypes with valuable genomic targets for \nunderstanding the biological mechanisms linking the diseases. \n \nIntroduction \nEndometriosis is a chronic gynaecological disease affecting up to 12% of reproductive-age \nwomen (Giudice, 2010; Rowlands et al, 2021; Treloar et al, 2002). The disease is \ncharacterised by the presence of endometriotic lesions outside the uterus and is associated \nwith pelvic pain and subfertility (Giudice, 2010).  Lesions are frequently categorized \naccording to lesion location and depth of infiltration into the surrounding tissue and include \nsuperficial peritoneal lesions, deep infiltrating disease and cysts (endometriomas), most \ncommonly found on the ovary (American Society for Reproductive, 1997). While \nendometriosis is a benign condition, it shares features with cancer including metastatic-like \nbehaviour, tissue invasion, proliferation, angiogenesis and decreased apoptosis. Large \nepidemiologic studies have reproducibly shown that women with endometriosis have \nincreased risk of epithelial ovarian cancer (standardized incidence ratio=1.8-8.95), although \nthe absolute risks are quite small, and there is currently no way to predict which \nendometriosis patients are most likely to develop ovarian cancer later in life (Brilhante et al, \n2017; Kok et al, 2015; Zafrakas et al, 2014). Somatic mutations shared between benign \nendometriotic lesions and adjacent tumours suggest these lesions are cellular precursors to \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nendometriosis-associated ovarian cancers. Examples include loss of function mutations in \nARID1A and gain of function mutations in PIK3CA (Anglesio et al, 2015; Jones et al, 2010; \nWiegand et al, 2010b, a).  \n \nOvarian cancer is the deadliest gynaecologic cancer. Fewer than 50% of women survive \nbeyond 5 years after diagnosis due to the rapid development of chemoresistance and the \nabsence of effective early detection strategies. Research is needed to advance understanding \nof disease aetiology, identify risk factors and develop early detection methods and effective \ntargeted therapies. The major histological subtypes of EOC include high-grade serous \n(HGSOC), low-grade serous (LGSOC), mucinous (MOC), endometrioid (ENOC) and clear \ncell (CCOC) EOC. Borderline tumors of low malignant potential also exist, most typically \nwith serous (LMPSOC) or mucinous differentiation. CCOC and ENOC are the two histotypes \nmost strongly and reproducibly associated with endometriosis (Brinton et al, 2005; Pearce et \nal, 2012; Rossing et al, 2008). Concurrent endometriosis is observed in 21-51% patients with \nCCOC and 23-43% women with ENOC (Jimbo et al, 1997; Stamp et al, 2016; Vercellini et \nal, 1993).  These histotype-associations are supported by observational study data from 7,911 \nwomen with invasive EOC in the Ovarian Cancer Association Consortium that showed a \nsignificant association between history of endometriosis and specific histological subtypes of \novarian cancer including CCOC (odds ratio (OR)=3.05), ENOC (OR=2.04) and, to a lesser \nextent, LGSOC (OR=2.11)(Pearce et al., 2012). \n \nRecent genome-wide association studies (GWASs) have provided strong evidence for a \ngenetic contribution to risk of both endometriosis and ovarian cancer. A total of 19 \nindependent genomic regions have been associated with endometriosis risk and 34 have been \nassociated with different histotypes of EOC (Phelan et al, 2017; Sapkota et al, 2017). \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nGermline variants may also contribute to increased risk of developing both diseases. Lu et al \n(2015) used genetic data from 84,000 SNPs genotyped in EOC (10,065 cases) and \nendometriosis (3,194 cases) cohorts to estimate the genetic correlation between the diseases \nand found a strong significant genetic correlation between endometriosis and CCOC (0.51), \nENOC (0.48), and LGSOC (0.40) and smaller correlation with HGSOC (0.25). This study, \nhowever, was limited by the relatively small number of SNPs and sample numbers and did \nnot provide evidence for a causal relationship between genetic liability to endometriosis and \nEOC risk or shared risk loci. The aim of this present study was to use data from newer and \nlarger endometriosis (14,949 cases/190,715 controls) and ovarian cancer (25,509 \ncases/40,941 controls) GWAS meta-analyses to estimate the genetic correlation and causal \nrelationship between endometriosis and ovarian cancer histotypes and to identify shared \ngenetic susceptibility loci, candidate target genes, and pathways. \n \nResults \nGenetic Correlation between Endometriosis and Ovarian Cancer Histotypes \nGenetic correlation can be used to describe the genetic relationship between two traits and is \nan estimate of the proportion of variance that two traits share that is attributed to genetics. \nEstimating the genetic correlation between traits contributes to our understanding of shared \nunderlying genetic risk factors and biological pathways. To estimate the genetic correlation \nbetween epithelial ovarian cancer (EOC) histotypes and endometriosis we used GWAS \nsummary statistics from meta-analyses conducted by Phelan et al. (2017) and  Sapkota et al. \n(2017) respectively, and linkage disequilibrium (LD) score regression (Bulik-Sullivan et al, \n2015). SNPs were matched on position and alleles to ensure effect size estimates were \nharmonised across data sets to obtain a set of 7,617,581 SNPs represented in the EOC \nhistotypes and endometriosis datasets. We estimated significant (P-value < 0.05) positive \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\ngenetic correlations (rg) between endometriosis and CCOC (rg=0.71), ENOC (rg=0.48) and \nHGSOC (rg=0.19) (Table 1). The rg for genetic correlation with LMPSOC was 0.88 but this \ndid not reach statistical significance, and we were unable to estimate rg for LGSOC due to \nthis histotype having the smallest sample size (1,012 cases). There was no evidence of a \nsignificant correlation between MOC and endometriosis.  \n \nMendelian Randomization Analysis \nWe then used Mendelian randomization based on the inverse-variance weighted (IVW) \nmethod (Burgess et al, 2013) and sensitivity analyses based on the weighted median \n(Bowden et al, 2016) and MR-Egger (Bowden et al, 2015) methods to investigate the \nassociation between genetic liability to endometriosis and EOC histotypes.  Genetic liability \nto endometriosis as predicted by 25 independent genome-wide significant (Rahmioglu et al, \n2018) (P < 5x10-8) endometriosis lead SNPs was associated with increased risk of CCOC, \nENOC, HGSOC and LMPSOC in the IVW analysis and the results were robust in sensitivity  \nanalyses (Table 2).  The strongest associations were observed for ENOC (P = 1.4x10-10; \nOR=1.66 (1.42-1.93)) and CCOC (P = 2.8x10-18, OR=2.59 (2.09-3.21)). \n \nGenetic Associations shared between Endometriosis and Ovarian Cancer Histotypes \nTo identify genetic associations with some evidence of a shared contribution from both \ndiseases, we combined the EOC histotypes and endometriosis susceptibility data sets using \ntwo complementary approaches; first, meta-analysis using approximate Bayes factors \ncomputed and combined by the MetABF method in both an independent and fixed model \n(Trochet et al, 2019) and second, meta-analysis based on the modified Han and Eskin \nrandom-effects model and fixed effects model implemented in RE2C (Han & Eskin, 2011; \nLee et al, 2017). The cross-trait meta-analysis identified several genome-wide significant \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nassociations and a summary of the number of SNPs nominally associated with both \nendometriosis and each ovarian cancer histotype using MetABF and RE2C are listed in Table \n3. SNPs were considered as markers of a shared genetic association with both traits if they \nhad (i) a log10 ABF > 4 in the cross-trait MetABF analysis using either model, (ii) a P-value < \n5x10-8 in the cross-trait RE2C analysis using either model, and (iii) a P-value < 0.05 in each \nsingle trait meta-analysis. A combined log10 ABF > 4 is equivalent to a posterior probability \nof combined association > 90% given a prior probability of association at any SNP of 1 in \n1,000. All SNPs (n=2,237 non-redundant) with P-value < 5x10-8 in RE2C had log10 ABF > 4 \nin MetABF suggesting good consistency between the methods. Filtering out SNPs that did \nnot have evidence for nominal association in each single trait meta-analysis (P < 0.05) \nfiltered out ~68% of the 3,612 SNPs, leaving 1,144 SNPs that met all three aforementioned \ncriteria. The largest number of shared genome-wide significant loci (or regions) were \nidentified between endometriosis and CCOC (14 loci; Tables 3 & 4). This was followed by \n13 risk loci shared between endometriosis and HGSOC, six risk loci with ENOC, five risk \nloci with MOC, five risk loci with LMPSOC and three risk loci with LGSOC (Tables 3 & 4). \nFour loci had lead SNPs with opposite directions of allelic association between endometriosis \nand the EOC histotype (Table 4). Significant SNPs in each analysis are listed in \nSupplementary Tables 1 & 2. Several loci also contain lead SNPs that have been associated \nwith other reproductive traits and diseases including uterine fibroids, sex hormone levels, \npolycystic ovarian syndrome (PCOS) and age at menarche (Supplementary Table 3). \n \nColocalization Analyses to identify Shared Causal Variants \nOur MetABF and RE2C analyses identified shared susceptibility loci for endometriosis and \nEOC. However, it is not clear whether the same candidate causal variants underlie the \nassociations at these loci or whether the associations at these loci are driven by distinct \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\ncandidate causal variants for endometriosis and EOC.  We examined the underlying shared \ngenetic architecture of endometriosis and EOC further using a statistical model to estimate \nthe posterior probability of association (PPA) that a genomic region 1) contains a variant \nassociated only with endometriosis (PPA_1), 2) contains a variant associated only with \novarian cancer (PPA_2), 3) contains a variant associated with both traits (PPA_3) and 4) \ncontains both a variant associated with endometriosis and an independent variant associated \nwith ovarian cancer (PPA_4). These models were implemented in gwas-pw (Pickrell et al, \n2016). Genomic regions with a PPA_3>0.5, evidence of a candidate causal variant \ninfluencing both diseases, or PPA_4>0.5, evidence that the candidate causal variants \nunderlying the association with each trait were distinct, are listed in Table 5. CCOC had the \nlargest number of genomic regions (n=13) with evidence of a shared causal variant with \nendometriosis. Figure 1 shows the genomic regions where there was statistical evidence for \nthe same causal variant underpinning endometriosis and at least one EOC histotype or where \nthere was evidence for two distinct signals. All regions identified with PPA_3 or PPA_4 >0.5 \ncontained lead SNPs significant in the cross-trait meta-analyses (log10 ABF > 4 in the cross-\ntrait MetABF analysis, P-value < 5x10-8 in the cross-trait RE2C analysis and P-value < 0.05 \nin each single trait meta-analysis) except for one region on chromosome 3 (Chr 3:126215130-\n128194265) where only colocalization offered evidence for a shared association between \nMOC and endometriosis.  Two regions with PPA_3>0.5 that achieved genome-wide \nsignificance in the meta-analyses (P-value < 5x10-8 and log10ABF > 4) were > 1 Mb from any \nrisk locus previously reported for endometriosis and EOC: 2q24.3 (rs13000026) and \n18p11.31 (rs10048393). One of the four loci (9p21) with lead SNPs with opposite directions \nof effect, identified in the cross-trait meta-analysis between endometriosis and CCOC, also \nhad evidence for the same causal variant underpinning both diseases from the colocalization \nanalysis. Another on chromosome 17 (17q12) had evidence for two distinct signals for \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nendometriosis and HGSOC. The remaining two had no evidence of colocalization. Several \ngenomic regions containing genome-wide significant associations identified in the cross-trait \nmeta-analyses only achieved PPA_1/2>0.5 suggesting the associations were only driven by \none of the two traits. However, this can also occur due to the limited power to detect \ncolocalization with the smaller sample sizes that were available for cross-trait colocalization \nanalyses involving the less common EOC histotypes. \n \nGene-based Association Analysis of Endometriosis and Ovarian Cancer Histotypes \nWe conducted a gene-based association analysis using fastBAT (Bakshi et al, 2016), a \nstatistical association test that calculates the combined association for all SNPs mapped to \neach gene while taking into account correlation between SNPs due to LD. Nine genes were \nassociated at genome-wide significance (P-value < 2.45x10-6) with endometriosis (GREB1, \nMIR4429, KDR, WNT4, SYNE1, CDKN2B-AS1, CDC42, ID4, PTPRO), 67 with HGSOC, \none with LGSOC (KIAA1024), four for LMPSOC (TERT, SLC6A18, MIR4457, CLPTM1L) \nand 27 for MOC in single trait gene-based analysis (Supplementary Table 4). Genome-wide \nsignificant genes for endometriosis were nominally significant (P < 0.05) for CCOC (GREB1, \nMIR4429, WNT4), ENOC (CDNK2B-AS1) and HGSOC (CDNK2B-AS1, MIR4429, WNT4) \n(Supplementary Table 4). \n \nWe looked at the overlap between the top 1% of genes associated with each trait (204/20,439 \ngenes evaluated in the fastBAT analysis) and observed an overlap of 5% between \nendometriosis and HGSOC (11 genes), 4% with CCOC (9 genes), 3% with LMPSOC (7 \ngenes), 3% with ENOC (6 genes), 3% with MOC (5 genes) and 1% with LGSOC (2 genes). \nTwo genes, SNX11 and CBX1, were associated with endometriosis, ENOC, HGSOC and \nLMPSOC. SKAP1 was associated with HGSOC, LMPSOC and endometriosis. However, \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nnone of the genes in the top 1% that overlapped between endometriosis and CCOC were in \nthe top 1% genes associated with other histotypes. Using an over-representation analysis in \nWebGestalt (Liao et al, 2019), no specific pathways were significantly enriched (FDR < \n0.05) for overlapping genes. This was also the case when the analysis was extended to the top \n5% of genes associated with each trait and the overlapping genes between endometriosis and \neach ovarian cancer histotype in the top 5% considered (Supplementary Table 4). \n \nFunctional Annotation \nWe collated all candidate causal variants by identifying all SNPs in tight linkage \ndisequilibrium (LD) with the lead SNPs (r2>0.7) from the cross-trait meta-analyses (log10 \nABF > 4 in the cross-trait MetABF analysis, P-value < 5x10-8 in the cross-trait RE2C \nanalysis and P-value < 0.05 in each single trait meta-analysis). The set of candidate causal \nvariants included 4,044 unique SNPs, which we functionally annotated to genes and \nepigenomic biofeatures: \n \nOverlap with noncoding DNA biofeatures \nTo identify putative functional SNPs, we overlapped all candidate causal SNPs with \nnoncoding regulatory elements (biofeatures) identified by epigenomic profiling of disease-\nrelevant tissues and cell lines. The biofeature catalogue consists of 11 consensus peak sets \n(see Methods, Supplementary Table 5) derived from 45 epigenomic profiles. Epigenome \nfeatures included open chromatin (18 ATAC-seq data sets) and active chromatin (27 \nH3K27ac ChIP-seq profiles; Supplementary Table 6). The specimens profiled include non-\ncancerous gynecologic tissues (fallopian tube, endometriosis and endometriosis-associated \nstroma) and ovarian cancer (clear cell, endometrioid, high-grade serous and mucinous) \ntissues or cell line models (Coetzee et al, 2015b; Corona et al, 2020a). Consensus peak sets \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\naveraged 33.6 (sd = 22, range = [9,84.3]) thousand peaks spanning, on average, 1.04% of the \nhuman genome (sd = 0.37, range = [0.42, 1.53]) (Supplementary Figure 1a-c, Supplementary \nTable 6). Genome coverage is marginally correlated with number of donors (Spearman’s rho \n= 0.43, p-value = 0.18; Supplementary Figure 1d-f). \n \nWe reduced the 1,144 candidate SNPs to 824 non-redundant variants most strongly \nassociated with both endometriosis and EOC histotypes (log10 ABF > 4 in the cross-trait \nMetABF analysis, a P-value < 5x10-8 in the cross-trait RE2C analysis and a P-value < 0.05 in \neach individual trait meta-analysis). Of these 824 candidate causal variants, 119 (14.4%) \noverlapped at least one biofeature (Figure 2a, Supplementary Table 7). The proportion of \nindependent loci containing SNPs intersecting with biofeatures varied by EOC histotype, \nwith only 33.3% of loci associated with endometriosis plus LGSOC overlapping with at least \none biofeature, while 71.4% of endometriosis and CCOC loci overlap with one or more \nrelevant biofeatures (Table 8,  Figure 2b, Supplementary Figure 2). As expected, ATAC-seq \nconsensus peak sets provide different information compared to H3K27ac ChIP-seq peak sets. \nWe observed that H3K27ac ChIP-seq consensus peak sets for fallopian tube, endometriosis-\nassociated stroma and endometriosis primary tissues and ATAC-seq consensus peaks for \nCCOC and fallopian tube intersect a similar set of SNPs, possibly reflecting the \nepidemiologic links between these tissues and diseases (Supplementary Figure 3).  \n \nThe 119 SNPs that overlap at least one ‘consensus peak set’ are distributed across 28 distinct \nloci (Supplementary Table 7). Overlaps provide functional evidence that these SNPs in risk \nloci shared between endometriosis and EOC histotypes are located within regulatory regions. \nThe MLLT10 and FSHB loci contains the SNPs with the most functional evidence and \nhighest number of overlaps, rs4071559 and rs10828247 each overlap eleven biofeatures \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\n(Supplementary Table 7). The VEZT locus harbors the SNPs with the second highest number \nof overlaps, where rs6538618 overlaps ten biofeatures at the putative bidirectional \nVEZT/FGD6 promoter (Figure 2c). rs6538618 has additional functional evidence and has \nbeen associated with the expression of both VEZT and FGD6 in endometrium (Mortlock et \nal, 2020b), fibroblasts, artery and muscle tissue (Consortium et al, 2017) (Figure 2c). The \nSKAP1 and PAX8 contain the greatest number of SNPs overlapping biofeatures (26 SNPs). \n \nTissue specific effects and disease relevant pathways \nUsing Functional Mapping and Annotation (FUMA)(Watanabe et al, 2017) we identified that \nthe expression of genes containing, or nearby to, SNPs shared between endometriosis and \ntwo ovarian cancer histotypes (CCOC & HGSOC) clustered across reproductive tissues \nincluding ovary, fallopian tube and uterus (Supplementary Figure 4). Several pathways were \nenriched within the set of genes annotated to significant SNPs (Supplementary Table 9). \nUnlike the fastBAT analysis genes were not identified using a gene-based association \nanalysis (SNPs within gene) but were instead annotated based on position (gene within 10kb \nof SNP). Focusing on enriched pathways containing three or more genes, pathways related to \ncell adhesion and nuclear division were enriched for genes annotated to SNPs associated with \nboth endometriosis and CCOC. Gene sets associated with other reproductive traits and \ndiseases were also enriched including uterine fibroids, endometrial cancer, dysmenorrheic \npain severity and gestational age at birth (Supplementary Table 9). \n \nCausal associations with gene expression and methylation \nThe fastBAT analysis involved a purely statistical gene-level association test. To complement \nfastBAT, we used Summary-data-based Mendelian Randomization (SMR)(Zhu et al, 2016) \nthat integrates gene-level expression and methylation with the GWAS data to elucidate \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\npotential gene-level functional mechanisms. SMR enables the identification of potentially \ncausal associations between shared susceptibility to endometriosis and ovarian cancer \nhistotypes and gene expression using SNPs associated with the traits from their individual \nGWA meta-analyses. We performed an SMR analysis using summary statistics from the \nendometriosis GWAS meta-analysis, each of the ovarian cancer histotype and eQTL data \nfrom endometrium (Fung et al, 2017; Mortlock et al, 2020a), blood (Võsa et al, 2018) and \nGTEx uterus and ovary (Consortium et al., 2017) (Supplementary Table 10). Associations \n(SMR P-value < 0.05; Heterogeneity In Dependent Instruments (HEIDI) P-value > 0.05) \nwere identified between variants, risk of endometriosis and at least one EOC histotype and \nexpression (P-value < 5x10-8) of 18 genes in endometrium, 25 genes in the ovary and 10 \ngenes in the uterus, the majority of genes only shared with a single histotype. This was not \ndependent on the same variant effecting the gene in both diseases. Variants were associated \nwith risk of endometriosis, CCOC, ENOC and expression of HLA-C in both the ovary and \nendometrium and risk of endometriosis and HGSOC and expression of HLA-K in ovary and \nuterus. \n \nPrevious studies have shown a large proportion of eQTLs are shared between tissues \n(Consortium et al., 2017; Mortlock et al., 2020b). To increase power the analysis was \nrepeated using a large blood cis-eQTL dataset from eQTLGen (Võsa et al., 2018) (n=31,684 \nindividuals) as a proxy and expression of 244 genes were associated (SMR P-value < 0.05; \nHEIDI P-value > 0.05) with both risk of endometriosis and ovarian cancer histotypes \n(Supplementary Table 10). Some of the top associations based on the blood eQTL data \ninclude expression of SKAP1 and risk of endometriosis, LMPSOC, CCOC and LGSOC, \nexpression of TNPO3 and IRF5 and risk of endometriosis, HGSOC and LMPSOC and \nexpression of CEP97 and risk of endometriosis and CCOC. \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\n \nSignificant associations (SMR P-value < 0.05; HEIDI P-value > 0.05) between variants, \nmethylation in the endometrium, and risk of endometriosis and at least one EOC histotype \nwere identified at 112 CpG sites including those near the GREB1 signal for endometriosis, \nCCOC, ENOC, HGSOC and MOC (Supplementary Table 10). Using a large blood \nmethylation QTL dataset (McRae et al, 2018) for SMR analysis, we identified variants \naffecting methylation at 1,449 CpG sites where variants associated with methylation were \nalso associated with endometriosis and at least one EOC histotype including sites near ESR1 \nfor CCOC, ENOC and MOC, SYNE1 for ENOC, HGSOC and LMPSOC, SKAP1 for CCOC, \nENOC and LGSOC and MLLT10 for CCOC, HGSOC, LGSOC and LMPSOC. Table 7 \nsummarises the various levels of evidence gained from the aforementioned analyses for loci \nassociated with both endometriosis and EOC histotypes. \n \nDiscussion \nAnalysis of germline genetic risk association data from endometriosis and ovarian cancer \nGWAS meta-analyses provides evidence of a genetic correlation and causal relationship \nbetween endometriosis and CCOC, ENOC and, to a lesser extent, HGSOC. Our results \nsupport epidemiological observations of an association between endometriosis and ovarian \ncancer as shown by estimates that women with endometriosis have two to three times higher \nrisk of developing ovarian cancer (Pearce et al., 2012; Wei et al, 2011) and that a high \nproportion of CCOC and ENOC cases also have endometriosis (20-50%)(Jimbo et al., 1997; \nStamp et al., 2016; Vercellini et al., 1993). Historically epidemiological studies have found \nlittle evidence for an association between endometriosis and HGSOC. However, a 2015 \nanalysis that involved evaluating genetic loci known at the time to be associated with \nendometriosis risk in a smaller subset of the Ovarian Cancer Association Consortium case-\n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\ncontrol set used here found significant evidence using a gene-based statistical test of an \nassociation with both endometriosis and HGSOC risk at the 1p36 (WNT4) locus (Lee et al, \n2016). Our Mendelian randomization (MR) results are consistent with findings from \nYarmolinsky et al (2019). However, potentially due to our use of a larger number of SNPs to \ninstrument endometriosis (25 SNPs based on the latest endometriosis GWAS versus 10 SNPs \nin the previously published analysis), we find that point estimates of the effect size for all \nassociations in our analysis are larger than those reported in Yarmolinsky, et al. This is most \nnotable in the odds ratio for CCOC (2.6 versus 1.5). The associations presented here reinforce \nthe well-established links between endometriosis and endometrioid and clear cell EOCs and \nwe also observe an association between endometriosis and HGSOC that was maintained \nacross more than one analytic approach, suggesting some shared pathways underlie the \ndevelopment of these two phenotypes. \n \nUsing cross-trait meta-analyses we identify 28 distinct genomic loci that shared a lead variant \ncontributing to the risk of both endometriosis and ovarian cancer histotypes. Colocalization \nanalyses provided evidence (PPA_3>0.5) for a single causal variant underlying risk for both \nendometriosis and ovarian cancer in 19 of these regions. Functional annotation revealed that \n14 of these 19 loci also contained risk SNPs that overlapped active and/or open chromatin. \nThe high posterior probability of colocalization at a large number of distinct loci is a \nremarkable feature of the genetic relationship between endometriosis and EOC histotypes and \nsuggests that identifying target genes in these loci may be valuable to understand the link \nbetween endometriosis and ovarian cancer and to intervene in neoplastic transformation. \n \nCombining GWAS data for conditions known to predispose to cancers and for the \ncorresponding cancers themselves has previously helped identify novel susceptibility loci for \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nnevus density and melanoma (Duffy et al, 2018) and for gastroesophageal reflux disease and \nesophageal cancer (An et al, 2019). Our analysis identified two risk loci that are completely \nnovel both in the context of endometriosis and EOC histotypes (i.e., > 1 Mb away from any \npreviously identified locus). At the first locus, located at chromosome2q24.3, the index SNP \n(rs13000026) lies intronic to Cordon-bleu protein-like 1 (COBLL1) and at the second locus at \n18p11.31, index SNP rs10048393 lies intronic to a long noncoding RNA, GAPLINC.   \nWhether these genes prove to be the target genes of these associations has yet to be \ndetermined; neither gene has been implicated in endometriosis nor ovarian cancer to date. \nThe lead SNPs in these regions displayed strong associations (P-value <= 3.2x10-4) with \nendometriosis risk and HGSOC risk in the single trait GWAS data sets and the combined \nsignal achieved genome-wide significance (P-value < 5x10-8). Moreover, gwas-pw \ncolocalization analysis of each these new loci indicated a high probability (>= 0.76) of a \nsingle causal signal underlying the association with both traits. \n \nDifferent regions shared between endometriosis and different histotypes may suggest \npossible biological mechanisms driving these causal relationships and the pathways \ncontributing to risk of specific subtypes. Three regions identified as associated with both \nCCOC and endometriosis using bivariate meta-analysis and gwas-pw, chromosome 4 near \nKDR, chromosome 8 near GDAP1 and chromosome 12 near VEZT, were not identified for \nENOC and HGSOC. Similarly, genomic regions on chromosome 1 near DNM3 and \nchromosome 11 near FSHB were associated with ENOC not CCOC or HGSOC, loci on \nchromosome 9 near ABO and chromosome 10 near MLLT10 were associated with HGSOC \nnot CCOC or ENOC. Alternatively, risk variants in the SKAP1 locus on chromosome 17 \nwere common between endometriosis and most histotypes. Shared variants in regions of \nknown hormone responsive genes, oestrogen-responsive growth regulation by oestrogen in \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nbreast cancer 1 (GREB1)(Mohammed et al, 2013) and kinase insert domain receptor \n(KDR)(Sugino et al, 2002), may suggest a role of hormone regulation in the causal pathway \nbetween endometriosis and CCOC. Cell adhesion pathways were also significantly enriched \nfor genes annotated to SNPs associated with risk of endometriosis and CCOC suggesting that \nthe ability of cells to adhere may contribute to the pathogenesis of endometriosis and \nsubsequently CCOC. Association between variants in the risk loci shared between \nendometriosis and EOC histotypes and other reproductive traits and diseases including \nPCOS, uterine fibroids and sex hormone levels suggests that perturbation of underlying \npathways important for the development and regulation of the reproductive and endocrine \nsystems may predispose women to a variety of diseases, the development of a particular \ndisease dependent on the presence of additional genetic and environmental risk factors.  \n \nInterestingly, the direction of effect at some shared risk loci differed between EOC \nhistotypes. The hepatocyte nuclear factor 1 beta (HNF1B) locus showed the same direction of \neffect between endometriosis and CCOC but was different between endometriosis and \nHGSOC consistent with published observations between CCOC and HGSOC (Natanzon et \nal, 2018). HNF1B is consistently highly expressed in CCOC but the promoter is methylated \nin HGSOC suggesting absence of HNF1B is critical for development of the HGSOC \nhistotype (Natanzon et al., 2018; Ross-Adams et al, 2016), potentially against a common \nbackground of genetic liability to endometriosis. HNF1B is a transcription factor that plays a \nvital role in tissue development, and regulation of genes involved in cell cycle modulation, \napoptosis, oxidative stress response and epithelial mesenchymal transition and dysregulation \nof these pathways may suggest a role for the microenvironment in tumor development \n(Suzuki et al, 2015; Yamaguchi et al, 2014). Similarly, SNPs in the SYNE1 locus on \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nchromosome 6 have the same direction of effect between endometriosis and CCOC but \nopposite direction of effect between endometriosis and LGSOC. \n \nWe provide evidence of functional mechanisms by which genetic variants associated with \nthese diseases may be impacting noncoding regulatory elements that control the expression of \ngenes that, when perturbed, increase risk of endometriosis and or ovarian cancer. Overall, \nmany target genes shared between endometriosis and EOC differed between histotypes \nsupporting evidence from other analyses in this study that different genes and gene pathways \nmay contribute to the causal relationship between endometriosis and the different histotypes. \nLINC00339, located in the chromosome 1 risk region was associated with risk of \nendometriosis and HGSOC. The expression of LINC00339 and nearby CDC42 has been \nassociated with endometriosis previously and LINC00339 has been reported as the likely \ntarget gene (Fung et al, 2018; Powell et al, 2016). Masuda et al (2019) also reports that this \nsame locus on chromosome 1 is associated with risk of both endometriosis and ovarian \ncancer. Methylation at a CpG site near GREB1 in endometrium and blood is associated with \nincreased risk for endometriosis, CCOC, ENOC, HGSOC and MOC. This association has \nbeen identified for endometriosis previously with functional studies yet to determine the \nmolecular mechanisms contributing to disease risk (Fung et al, 2015; Mortlock et al, 2019). \nTranscription of GREB1 splice variants has been associated with variants in this region in \novarian tissue (2020). This gene is expressed in EOC tumours with studies suggesting a \nreliance on ESR1/GREB1 signalling (Hodgkinson et al, 2018; Laviolette et al, 2014).  \n \nOverlap with chromatin biofeatures in ovarian and endometriosis tissues also highlighted \npotential target genes. Risk SNPs in the VEZT region overlapped the putative bidirectional \npromoter for VEZT and FGD6. The lead SNP from the bi-variate meta-analysis in the VEZT \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nlocus (rs7309252) was in LD (r2=0.99) with a SNP (rs6538618) overlapping 10 regulatory \nbiofeatures. Expression of both VEZT and FGD6 have been associated with endometriosis \nrisk previously(Fung et al., 2018; Mortlock et al., 2020b). Similarly, the lead SNPs in the \nFSHB and MLLT10 loci were in LD (r2>0.8) with SNPs overlapping 11 biofeatures in the \npromoter region of ADP Ribosylation Factor Like GTPase 14 Effector Protein (ARL14EP) \nand MLLT10 respectively and were associated with methylation at nearby CpG sites. Risk \nSNPs associated with endometriosis, HGSOC and LMPSOC span follicle stimulating \nhormone (FSH) subunit B (FSHB) and nearby ARL14EP. ARL14EP is expressed by many \ntissue types and plays a role in the movement of major histocompatibility class II molecules \nalong the actin cytoskeleton. FSHB is expressed in the pituitary gland and plays an important \nrole regulating reproductive function. Variants in the 11p14.1 locus near FSHB have been \nsignificantly associated with multiple reproductive traits and diseases including PCOS, \nuterine fibroids, circulating sex hormone levels and menstrual cycle characteristics (Day et \nal, 2018; Day et al, 2015; Day et al, 2017; Gallagher et al, 2019; Laisk et al, 2018; Mbarek et \nal, 2016; Ruth et al, 2016; Sapkota et al., 2017). The lead SNP from the bi-variate meta-\nanalysis, rs11031005, is in LD with a FSHB promoter polymorphism (rs10835638), and \nenhancer polymorphism (rs11031006) involved in regulating the FSHB gene transcription \n(Bohaczuk et al, 2021; Grigorova et al, 2008; Trevisan et al, 2019). The locus containing the \nhistone lysine methyltransferase DOT1L Cofactor (MLLT10) was associated with \nendometriosis risk in a recent endometriosis GWAS (Rahmioglu et al., 2018). Studies have \nalso linked ovarian cancer susceptibility and endometriosis risk to subtle variations in \nregulation at the MLLT10 promoter region. SNPs in LD (r2>0.8) with the lead variant from \nthe bivariate meta-analysis have been annotated to the promoter of MLLT10 and have been \nassociated with changes of expression of nearby genes C10orf140, C10orf114 and NEBL in  \nprimary EOC tissues and changes in expression of NEBL in endometrium, suggesting this \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\npromoter may also have cis regulatory activity across the locus (Mortlock et al., 2020b; \nPharoah et al, 2013). \n \nThis study uses a comprehensive range of statistical genetic approaches to build on existing \nevidence of an association between endometriosis and ovarian cancer using genetic data from \nthe largest GWAS meta-analyses of endometriosis and EOC currently available. The power \nof this study to identify shared risk loci and target genes is however limited by the sample \nsize of some of the less common EOC histotype cohorts such as LMPSOC. The identification \nof genetic relationships may also be limited by phenotypic annotation and heterogeneity \nbetween endometriosis cases affecting the endometriosis GWAS. Several studies have \nreported an association between endometriosis sub-phenotypes and risk of ovarian cancer, in \nparticular endometriomas (Buis et al, 2013; Kobayashi et al, 2007; Kok et al., 2015; \nSaavalainen et al, 2018). More comprehensive phenotyping and molecular characterisation of \nendometriosis lesions could be used to test genetic associations between potential \nendometriosis subtypes and risk of certain EOC histotypes, for example testing if the \nassociation between endometriosis and CCOC is driven by endometriomas specifically.  \n \nIn conclusion we find evidence of a strong genetic correlation and causal relationship \nbetween endometriosis and two ovarian cancer histotypes CCOC and ENOC and to a lesser \nextent HGSOC. Further investigation into shared genomic regions reveals different genetic \nvariants, genes and pathways are likely contributing to the causal relationship with the \ndifferent histotypes. Results add to our understanding of disease pathogenesis and yield \ngenomic targets that may facilitate preventive pharmacological intervention by disrupting the \nlink between endometriosis and ovarian cancer and promote targeted ovarian cancer \nscreening in women with endometriosis. \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\n \nMethods \nDatasets \nEpithelial Ovarian Cancer \nGenetic data in the form of GWAS summary statistics were available from a 2017 GWAS \nmeta-analysis for EOC and EOC histotypes conducted by Phelan et al. (2017). Summary \nstatistics included the SNP RSID, effect allele, other allele, effect allele frequency, beta \ncoefficient (be), standard error (SE) and p-value. The meta-analysis included 25,509 EOC \ncases and 40,941 controls of European ancestry with association statistics reported for \n10,197,379 SNPs for overall EOC, high-grade serous ovarian cancer (HGSOC), low-grade \nserous ovarian cancer (LGSOC), low malignant potential serous ovarian cancer (LMPSOC), \nmucinous ovarian cancer (MOC), endometrioid ovarian cancer (ENOC) and clear cell \novarian cancer (CCOC). The number of cases and controls included in each histotype \nanalysis are shown in Supplementary Table 11. Summary statistics were filtered to remove \nSNPs with an imputation quality r2 (OncoArray) score < 0.3 and SNPs with a minor allele \nfrequency (MAF) < 0.01 leaving 10,197,379 SNPs for subsequent analysis. \n \nEndometriosis \nEndometriosis GWAS summary statistics were available from the 2017 GWAS meta-analysis \nfor endometriosis conducted by Sapkota et al. (2017). Summary statistics include the SNP \nRSID, effect allele, other allele, beta coefficient (be), standard error (SE), effect allele \nfrequency, and p-value. Only statistics generated from European cohorts was used in \nsubsequent analyses including 14,949 cases and 190,715 controls. Summary statistics for \n7,899,415 SNPs remained following removal of imputed genotypes with low imputation \nquality (<0.3 for minimac and <0.4 for IMPUTE2) and SNPs with a MAF < 0.01. \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\n \nGenetic Correlation and Mendelian Randomization \nLinkage disequilibrium score regression (LDSC) was used to estimate the genetic correlation \nbetween endometriosis and ovarian cancer histotypes using the GWAS summary statistics \n(Bulik-Sullivan et al., 2015). LD scores were computed using 1000 Genomes European \nancestry data as the independent variable in the LD Score regression and for the regression \nweights. In the absence of sample overlap between the datasets we also constrained the LD \nscore regression intercept to reduce the standard error substantially. Mendelian randomization \n(MR) analyses were performed using the R package MendelianRandomization version 0.5.0 \n(Yavorska & Burgess, 2017).  Inverse-variance weighted (IVW)(Burgess et al., 2013) \nregression was used for primary analysis and weighted median (Bowden et al., 2016) and \nMR-Egger regression(Bowden et al., 2015) for sensitivity analyses.  Associations were \ndeclared significant if both IVW and weighted median analyses yield P-values < 0.05 and the \ndirection of the effect size estimates (odds ratios) were consistent across IVW, weighted \nmedian, MR-Egger regression approaches.  Of the 27 independent, genome-wide significant \nendometriosis lead risk SNPs (Rahmioglu et al., 2018), two multiallelic SNPs (rs484686 and \nrs4762173) were removed, leaving 25 SNPs in the instrument for genetic liability to \nendometriosis. \n \nSNP-based Association Analysis \nCross-trait Meta-analysis \nTo identify risk loci associated with both traits we conducted a bivariate meta-analysis using \ntwo different approaches, MetABF (Trochet et al., 2019) and Han and Eskin random-effects \nmodel (RE2C) (Han & Eskin, 2011; Lee et al., 2017). MetABF is a method to meta-analyse \ngenome-wide association studies using approximate Bayes Factors. Beta coefficients (effect \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nsizes) and SEs from the endometriosis dataset were used as input for MetABF alongside beta \ncoefficients and SEs from ovarian cancer histotypes. Each of the six ovarian cancer \nhistotypes were meta-analysed with endometriosis separately. Correlation in effect sizes was \nmodeled using both an independent and fixed effect model and the prior parameter for the \nvariance in effect sizes, sigma, was set to 0.1 allowing effect sizes to be small, characteristic \nof complex diseases. Significantly associated SNPs were defined by log10 ABF > 4 in either \nthe fixed or independent model and at least nominal significance in the single trait GWAS (P-\nvalue<0.05). \n \nTo validate results, we also performed a routine fixed effects meta-analysis and a modified \nrandom effects meta-analysis on the same data using RE2C. RE2C is designed to integrate \nthe effects while accounting for the heterogeneity between studies. A SNP was deemed \nsignificantly associated with both traits in the bivariate meta-analysis if it met a fixed \n(Cochran's Q statistic P-value >0.05) or random (Cochran's Q statistics's P-value <0.05) \neffect threshold of P-value < 5x10-8, and was at least nominally significant in the single trait \nGWAS meta-analysis (P-value<0.05) and had no significant heterogeneity. SNPs found to be \nsignificantly associated with both traits and passing thresholds for both Metabf and RE2C \nwere fine mapped using FUMA (Watanabe et al., 2017) to identify independent signals to \nidentify independent signals at r2 < 0.6. \n \nColocalization Analyses \nGWAS-pw (Pickrell et al., 2016) was used to estimate the probability that in a given genomic \nregion the same variant underlies the association with both traits. The genome is split into \n1703 non-overlapping regions using linkage disequilibrium blocks (Berisa & Pickrell, 2015) \nand following Giambartolomei et al (2014) the software estimates the probability, using an \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nempirical Bayes approach, that a given genomic region either 1) contains a genetic variant \nthat influences the first trait, 2) contains a genetic variant that influences the second trait, 3) \ncontains a genetic variant that influences both traits (posterior probability of association, \nPPA3), or 4) contains both a genetic variant that influences the first trait and a separate \ngenetic variant that influences the second trait (PPA4). This was performed using the \nendometriosis dataset and each ovarian cancer histotype. Regions with a PPA3/4>0.5 were \nconsidered to have evidence of a shared causal variant and independent causal variants \nrespectively. \n \nGene-based Association Analysis \nThe fastBat function in GCTA (Bakshi et al., 2016) was used to perform a gene-based \nassociation analysis for endometriosis and each ovarian cancer histotype using GWAS \nsummary statistics from each trait. A total of 20,439 genes (hg19) were tested for each \ndisease using an LD cutoff of 0.9 and no SNPs outside defined gene boundaries. Genes in the \ntop 1% and 5% associated with each EOC histotype that overlap with the top 1% and 5% of \ngenes associated with endometriosis were tested for enrichment in pathways using over-\nrepresentation analysis in WebGestalt (Liao et al., 2019). \n \nFunctional Annotation \nThe GENE2FUNC option in FUMA (Watanabe et al., 2017) was also used to investigate the \nexpression of nearby (within 10kb) genes across tissues and enrichment of functional and \nbiological pathways. \n \nH3K27ac ChIP-seq and ATAC-seq \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nH3K27ac ChIP-seq data have been previously described (Coetzee et al, 2015a; Corona et al, \n2020b; Jones et al, 2020) with the exception of the endometrioma and endometrioma stroma \nspecimens, which were profiled in parallel with the tumors reported in Corona et al. (2020b). \n \nATAC-seq data were generated for primary tissues collected, with informed consent, as part \nof the Gynecologic Tissue Bank at Cedars-Sinai Medical Center. Tumors were OCT \nembedded and H&Es reviewed to confirm the diagnosis. Epithelial-rich regions of flash \nfrozen tumors were biopsied using a 5mm biopsy punch. Primary fallopian tube tissues were \nsubjected to enzymatic digest, as previously described (Fotheringham et al, 2011; Karst & \nDrapkin, 2012) and the epithelial-enriched cells frozen down for ATAC-seq. In one instance, \nfallopian epithelial cells were frozen prior to profiling. ATAC-seq was performed by Active \nMotif. \n \nPeak calling - Peak calling of H3K27ac ChIP-seq profiles was performed using the \nENCODE pipeline (v1.2.2) with p-val_thresh = 1e-09, reference genome hg38 and other \ndefault parameters. If two technical replicates were used, we selected the overlap peak set of \nthe true replicates (rep1_rep2.overlap.bfilt.narrowPeak.gz) as the ‘sample peak set’, \notherwise, we used the overlap peak set of the pseudo replicates (rep1-\npr.overlap.bfilt.narrowPeak.gz). Quality control metrics (number of unique mapped reads, \ncross-correlation metrics, IDR) were checked for each technical replicate (Supplementary \nTable 12). \n \nDonor peak set - Two or more sample peak sets are combined if they belong to the same \ndonor, resulting in a set of non-overlapping peaks. First, individual peak scores (-log10(p-\nvalue)) are normalized to a ‘score per million’ (SPM) dividing the peak score, by the sum of \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nall peak scores within the sample, divided by a factor of a million. Then, all peaks from the \nsame donor are merged and sorted by SPM. Iteratively, a peak is removed if it overlaps \nanother peak with a higher SPM, resulting in a set of non-overlapping peaks that represent a \ndonor. \n \nConsensus peak set - A set of sample/donor peak sets will be combined when they are part of \nthe same experiment, tissue type and sample type triage. Similarly to the donor peak set \nprocedure, peaks are merged, sorted and iteratively removed, based on the SPM. Then, the \nremaining non-overlapping peaks are tagged as reproducible if they overlap peaks with an \nSPM > 5 in two or more samples. Non-reproducible peaks, peaks that are within repeat mask \nregions and peaks in the “Y” chromosome are removed, leaving a set of reproducible non-\noverlapping peaks as the ‘consensus peak set’. Consensus peak sets are then transformed \nfrom hg38 to hg19 using liftOver with parameteres -bedPlus=6, the \nhg38toHg19.over.chain.gz from UCSC chain files and other default parameters. On average, \neach consensus peak set has 125.2 (0.28%) genomic locations that could not be mapped \n(range = [8,350]). \n \nSummary-data-based Mendelian randomization (SMR) \nSummary-data-based Mendelian randomization (SMR) is used to assess the pleiotropic and \ncausal association between genetic variants, gene expression/methylation level and risk of \ndisease (Zhu et al., 2016). SMR has been conducted using summary statistics from Sapkota et \nal. (2017) and endometrial and blood eQTLs and mQTLs (Fung et al., 2017; McRae et al., \n2018; Mortlock et al., 2020a; Mortlock et al., 2019). Using these published results, we \nsearched for any significant SMR associations in loci associated (log10 ABF > 4 in the cross-\ntrait MetABF analysis, P-value < 5x10-8 in the cross-trait RE2C analysis and P-value < 0.05 \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nin each single trait meta-analysis) with both endometriosis and ovarian cancer from the \nbivariate meta-analysis. We also conducted SMR on each ovarian cancer histotype by \nintegrating the GWAS meta-analysis summary data from Phelan et al. (2017) and summary \neQTL data from endometrium (Fung et al., 2017; Mortlock et al., 2020a), eQTLGen (Võsa et \nal., 2018) and GTEx ovary and uterus (Consortium et al., 2017). Associations were \nconsidered significant if they had a PSMR<0.05/(number of genes tested) and a PHEIDI > \n0.05/(number of genes passing the SMR test). Of note, associations with PSMR<0.05 and a \nPHEIDI > 0.05 were also considered due to power limitations in existing datasets. In the \nabsence of multiple testing correction these would require future validation in larger datasets \nor functional studies. The heterogeneity in dependent instruments (HEIDI) test considers the \npattern of risk associations using all the SNPs that are significantly associated with gene \nexpression in a region and evaluates the null hypothesis that there is a single variant affecting \ngene expression/methylation and disease risk and the alternative hypothesis that there are \ndistinct variants associated with expression/methylation and disease. \n \nData Availability  \nThis study includes no data deposited in external repositories. \n \nAcknowledgements \nWe would like to thank the research participants and employees of 23andMe for making this \nwork possible. We would also like to acknowledge the International Endometriosis Genetics \nConsortium and Ovarian Cancer Association Consortium for their contributions generating \nthe GWAS datasets and data access. We are also grateful to the thousands of patients who \ndonated the specimens that enable this research to happen. For acknowledgements for the \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nendometriosis meta-analysis please see Sapkota et al. (2017). For acknowledgements for \nthe ovarian cancer meta-analysis please see Phelan et al. (2017). \n \nFunding \nThis work was supported by the National Health and Medical Research Council of Australia \n(GNT1026033, GNT1105321, GNT1147846, Investigator Grant 1177194 to G.W.M and \nMedical Research Future Fund Research Grant MRF1199785 to S.M) and National Institutes \nof Health (R01CA193910, R01CA204954, R01CA211707, R01CA251555). S.P.K is \nsupported by a United Kingdom Research and Innovation Future Leaders Fellowship \n(MR/T043202/1). K.L. is supported by a Liz Tilberis Early Career Award (599175) and a \nProgram Project Development (373356) from the Ovarian Cancer Research Alliance, plus a \nResearch Scholar’s Grant from the American Society (134005). For funding details of the \nendometriosis meta-analysis please see Sapkota et al. (2017). For funding details of \nthe ovarian cancer meta-analysis please see Phelan et al. (2017). \n \nAuthor contributions \nS.M, S.P.K, K.L, G.W.M and P.P designed the study with input from the other authors. Data \nanalysed in this study was generated by the Ovarian Cancer Association Consortium and \nInternational Endometriosis Genetics Consortium. S.M, and S.P.K ran additional quality \ncontrol and filtering of GWAS datasets. J-H.S, M.L.F, S.A.G, M.T.S, R.L, C.W, I.C, B.Y.K, \nP.A.W.R and B.J.R contributed to specimen collection and data generation. Data analysis \nwas performed by S.M, S.P.K, R.I.C and P.F.K which was interpreted by all authors. S.M, \nS.P.K and K.L drafted the report with input from all other authors. The final manuscript has \nbeen critically revised and approved by all authors. \n \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nConflict of interest \nThe authors declare that they have no conflict of interest. \n \nReferences \n(2020) The GTEx Consortium atlas of genetic regulatory effects across human tissues. \nScience 369: 1318 \nAmerican Society for Reproductive M (1997) Revised American Society for Reproductive \nMedicine classification of endometriosis: 1996. Fertility and Sterility 67: 817-821 \nAn J, Gharahkhani P, Law MH, Ong JS, Han X, Olsen CM, Neale RE, Lai J, Vaughan TL, \nGockel I et al (2019) Gastroesophageal reflux GWAS identifies risk loci that also associate \nwith subsequent severe esophageal diseases. Nat Commun 10: 4219 \nAnglesio MS, Bashashati A, Wang YK, Senz J, Ha G, Yang W, Aniba MR, Prentice LM, \nFarahani H, Li Chang H et al (2015) Multifocal endometriotic lesions associated with cancer \nare clonal and carry a high mutation burden. The Journal of pathology 236: 201-209 \nBakshi A, Zhu Z, Vinkhuyzen AAE, Hill WD, McRae AF, Visscher PM, Yang J (2016) Fast \nset-based association analysis using summary data from GWAS identifies novel gene loci for \nhuman complex traits. Scientific Reports 6: 32894 \nBerisa T, Pickrell JK (2015) Approximately independent linkage disequilibrium blocks in \nhuman populations. Bioinformatics 32: 283-285 \nBohaczuk SC, Thackray VG, Shen J, Skowronska-Krawczyk D, Mellon PL (2021) FSHB \n Transcription is Regulated by a Novel 5' Distal Enhancer With a Fertility-Associated Single \nNucleotide Polymorphism. Endocrinology 162 \nBowden J, Davey Smith G, Burgess S (2015) Mendelian randomization with invalid \ninstruments: effect estimation and bias detection through Egger regression. International \nJournal of Epidemiology 44: 512-525 \nBowden J, Davey Smith G, Haycock PC, Burgess S (2016) Consistent Estimation in \nMendelian Randomization with Some Invalid Instruments Using a Weighted Median \nEstimator. Genet Epidemiol 40: 304-314 \nBrilhante AVM, Augusto KL, Portela MC, Sucupira LCG, Oliveira LAF, Pouchaim AJMV, \nNóbrega LRM, Magalhães TFd, Sobreira LRP (2017) Endometriosis and Ovarian Cancer: an \nIntegrative Review (Endometriosis and Ovarian Cancer). Asian Pac J Cancer Prev 18: 11-16 \nBrinton LA, Sakoda LC, Sherman ME, Frederiksen K, Kjaer SK, Graubard BI, Olsen JH, \nMellemkjaer L (2005) Relationship of Benign Gynecologic Diseases to Subsequent Risk of \nOvarian and Uterine Tumors. Cancer Epidemiology Biomarkers &amp;amp; Prevention 14: \n2929 \nBuis CC, van Leeuwen FE, Mooij TM, Burger CW (2013) Increased risk for ovarian cancer \nand borderline ovarian tumours in subfertile women with endometriosis. Human \nreproduction (Oxford, England) 28: 3358-3369 \nBulik-Sullivan B, Finucane HK, Anttila V, Gusev A, Day FR, Loh P-R, Duncan L, Perry \nJRB, Patterson N, Robinson EB et al (2015) An atlas of genetic correlations across human \ndiseases and traits. Nature Genetics 47: 1236-1241 \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nBurgess S, Butterworth A, Thompson SG (2013) Mendelian Randomization Analysis With \nMultiple Genetic Variants Using Summarized Data. Genetic Epidemiology 37: 658-665 \nCoetzee SG, Shen HC, Hazelett DJ, Lawrenson K, Kuchenbaecker K, Tyrer J, Rhie SK, \nLevanon K, Karst A, Drapkin R et al (2015a) Cell-type-specific enrichment of risk-\nassociated regulatory elements at ovarian cancer susceptibility loci. Hum Mol Genet 24: \n3595-3607 \nCoetzee SG, Shen HC, Hazelett DJ, Lawrenson K, Kuchenbaecker K, Tyrer J, Rhie SK, \nLevanon K, Karst A, Drapkin R et al (2015b) Cell-type-specific enrichment of risk-\nassociated regulatory elements at ovarian cancer susceptibility loci. Human molecular \ngenetics 24: 3595-3607 \nConsortium G, Aguet F, Brown AA, Castel SE, Davis JR, He Y, Jo B, Mohammadi P, Park \nY, Parsana P et al (2017) Genetic effects on gene expression across human tissues. Nature \n550: 204 \nCorona RI, Seo J-H, Lin X, Hazelett DJ, Reddy J, Fonseca MAS, Abassi F, Lin YG, \nMhawech-Fauceglia PY, Shah SP et al (2020a) Non-coding somatic mutations converge on \nthe PAX8 pathway in ovarian cancer. Nature Communications 11: 2020 \nCorona RI, Seo JH, Lin X, Hazelett DJ, Reddy J, Fonseca MAS, Abassi F, Lin YG, \nMhawech-Fauceglia PY, Shah SP et al (2020b) Non-coding somatic mutations converge on \nthe PAX8 pathway in ovarian cancer. Nat Commun 11: 2020 \nDay F, Karaderi T, Jones MR, Meun C, He C, Drong A, Kraft P, Lin N, Huang H, Broer L et \nal (2018) Large-scale genome-wide meta-analysis of polycystic ovary syndrome suggests \nshared genetic architecture for different diagnosis criteria. PLOS Genetics 14: e1007813 \nDay FR, Ruth KS, Thompson DJ, Lunetta KL, Pervjakova N, Chasman DI, Stolk L, Finucane \nHK, Sulem P, Bulik-Sullivan B et al (2015) Large-scale genomic analyses link reproductive \naging to hypothalamic signaling, breast cancer susceptibility and BRCA1-mediated DNA \nrepair. Nature Genetics 47: 1294 \nDay FR, Thompson DJ, Helgason H, Chasman DI, Finucane H, Sulem P, Ruth KS, Whalen \nS, Sarkar AK, Albrecht E et al (2017) Genomic analyses identify hundreds of variants \nassociated with age at menarche and support a role for puberty timing in cancer risk. Nature \nGenetics 49: 834 \nDuffy DL, Zhu G, Li X, Sanna M, Iles MM, Jacobs LC, Evans DM, Yazar S, Beesley J, Law \nMH et al (2018) Novel pleiotropic risk loci for melanoma and nevus density implicate \nmultiple biological pathways. Nat Commun 9: 4774 \nFotheringham S, Levanon K, Drapkin R (2011) Ex vivo culture of primary human fallopian \ntube epithelial cells. J Vis Exp \nFung JN, Girling JE, Lukowski SW, Sapkota Y, Wallace L, Holdsworth-Carson SJ, Henders \nAK, Healey M, Rogers PAW, Powell JE et al (2017) The genetic regulation of transcription \nin human endometrial tissue. Human Reproduction 32: 893-904 \nFung JN, Holdsworth-Carson SJ, Sapkota Y, Zhao ZZ, Jones L, Girling JE, Paiva P, Healey \nM, Nyholt DR, Rogers PAW et al (2015) Functional evaluation of genetic variants associated \nwith endometriosis near GREB1. Human Reproduction 30: 1263-1275 \nFung JN, Mortlock S, Girling JE, Holdsworth-Carson SJ, Teh WT, Zhu Z, Lukowski SW, \nMcKinnon BD, McRae A, Yang J et al (2018) Genetic regulation of disease risk and \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nendometrial gene expression highlights potential target genes for endometriosis and \npolycystic ovarian syndrome. Scientific Reports 8: 11424 \nGallagher CS, Mäkinen N, Harris HR, Rahmioglu N, Uimari O, Cook JP, Shigesi N, Ferreira \nT, Velez-Edwards DR, Edwards TL et al (2019) Genome-wide association and \nepidemiological analyses reveal common genetic origins between uterine leiomyomata and \nendometriosis. Nature Communications 10: 4857 \nGiambartolomei C, Vukcevic D, Schadt EE, Franke L, Hingorani AD, Wallace C, Plagnol V \n(2014) Bayesian Test for Colocalisation between Pairs of Genetic Association Studies Using \nSummary Statistics. PLOS Genetics 10: e1004383 \nGiudice LC (2010) Clinical Practice: Endometriosis. The New England journal of medicine \n362: 2389-2398 \nGrigorova M, Punab M, Ausmees K, Laan M (2008) FSHB promoter polymorphism within \nevolutionary conserved element is associated with serum FSH level in men. Hum Reprod 23: \n2160-2166 \nHan B, Eskin E (2011) Random-Effects Model Aimed at Discovering Associations in Meta-\nAnalysis of Genome-wide Association Studies. The American Journal of Human Genetics \n88: 586-598 \nHodgkinson K, Forrest LA, Vuong N, Garson K, Djordjevic B, Vanderhyden BC (2018) \nGREB1 is an estrogen receptor-regulated tumour promoter that is frequently expressed in \novarian cancer. Oncogene 37: 5873-5886 \nJimbo H, Yoshikawa H, Onda T, Yasugi T, Sakamoto A, Taketani Y (1997) Prevalence of \novarian endometriosis in epithelial ovarian cancer. International Journal of Gynecology & \nObstetrics 59: 245-250 \nJones MR, Peng PC, Coetzee SG, Tyrer J, Reyes ALP, Corona RI, Davis B, Chen S, Dezem \nF, Seo JH et al (2020) Ovarian Cancer Risk Variants Are Enriched in Histotype-Specific \nEnhancers and Disrupt Transcription Factor Binding Sites. American Journal of Human \nGenetics 107: 622-635 \nJones S, Wang T-L, Shih I-M, Mao T-L, Nakayama K, Roden R, Glas R, Slamon D, Diaz \nLA, Vogelstein B et al (2010) Frequent Mutations of Chromatin Remodeling Gene ARID1A \nin Ovarian Clear Cell Carcinoma. Science 330: 228 \nKarst AM, Drapkin R (2012) Primary culture and immortalization of human fallopian tube \nsecretory epithelial cells. Nat Protoc 7: 1755-1764 \nKobayashi H, Sumimoto K, Moniwa N, Imai M, Takakura K, Kuromaki T, Morioka E, \nArisawa K, Terao T (2007) Risk of developing ovarian cancer among women with ovarian \nendometrioma: a cohort study in Shizuoka, Japan. Int J Gynecol Cancer 17: 37-43 \nKok VC, Tsai H-J, Su C-F, Lee C-K (2015) The Risks for Ovarian, Endometrial, Breast, \nColorectal, and Other Cancers in Women With Newly Diagnosed Endometriosis or \nAdenomyosis: A Population-Based Study. International Journal of Gynecologic Cancer 25: \n968 \nLaisk T, Kukuškina V, Palmer D, Laber S, Chen C-Y, Ferreira T, Rahmioglu N, Zondervan \nK, Becker C, Smoller JW et al (2018) Large-scale meta-analysis highlights the \nhypothalamic–pituitary–gonadal axis in the genetic regulation of menstrual cycle length. \nHuman molecular genetics 27: 4323-4332 \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nLaviolette LA, Hodgkinson KM, Minhas N, Perez-Iratxeta C, Vanderhyden BC (2014) 17β-\nestradiol upregulates GREB1 and accelerates ovarian tumor progression in vivo. \nInternational Journal of Cancer 135: 1072-1084 \nLee AW, Templeman C, Stram DA, Beesley J, Tyrer J, Berchuck A, Pharoah PP, Chenevix-\nTrench G, Pearce CL, Ovarian Cancer Association C (2016) Evidence of a genetic link \nbetween endometriosis and ovarian cancer. Fertility and sterility 105: 35-43.e31-10 \nLee CH, Eskin E, Han B (2017) Increasing the power of meta-analysis of genome-wide \nassociation studies to detect heterogeneous effects. Bioinformatics 33: i379-i388 \nLiao Y, Wang J, Jaehnig EJ, Shi Z, Zhang B (2019) WebGestalt 2019: gene set analysis \ntoolkit with revamped UIs and APIs. Nucleic Acids Research 47: W199-W205 \nLu Y, Cuellar-Partida G, Painter JN, Nyholt DR, Australian Ovarian Cancer S, International \nEndogene C, Morris AP, Fasching PA, Hein A, Burghaus S et al (2015) Shared genetics \nunderlying epidemiological association between endometriosis and ovarian cancer. Human \nmolecular genetics 24: 5955-5964 \nMasuda T, Low S-K, Akiyama M, Hirata M, Ueda Y, Matsuda K, Kimura T, Murakami Y, \nKubo M, Kamatani Y et al (2019) GWAS of five gynecologic diseases and cross-trait \nanalysis in Japanese. European Journal of Human Genetics \nMbarek H, Steinberg S, Nyholt DR, Gordon SD, Miller MB, McRae AF, Hottenga JJ, Day \nFR, Willemsen G, de Geus EJ et al (2016) Identification of Common Genetic Variants \nInfluencing Spontaneous Dizygotic Twinning and Female Fertility. American journal of \nhuman genetics 98: 898-908 \nMcRae AF, Marioni RE, Shah S, Yang J, Powell JE, Harris SE, Gibson J, Henders AK, \nBowdler L, Painter JN et al (2018) Identification of 55,000 Replicated DNA Methylation \nQTL. Scientific Reports 8: 17605 \nMohammed H, D’Santos C, Serandour Aurelien A, Ali HR, Brown Gordon D, Atkins A, \nRueda Oscar M, Holmes Kelly A, Theodorou V, Robinson Jessica LL et al (2013) \nEndogenous Purification Reveals GREB1 as a Key Estrogen Receptor Regulatory Factor. \nCell Reports 3: 342-349 \nMortlock S, Kendarsari RI, Fung JN, Gibson G, Yang F, Restuadi R, Girling JE, Holdsworth-\nCarson SJ, Teh WT, Lukowski SW et al (2020a) Tissue specific regulation of transcription in \nendometrium and association with disease. Human Reproduction 35: 377-393 \nMortlock S, Kendarsari RI, Fung JN, Gibson G, Yang F, Restuadi R, Girling JE, Holdsworth-\nCarson SJ, Teh WT, Lukowski SW et al (2020b) Tissue specific regulation of transcription in \nendometrium and association with disease. Human Reproduction \nMortlock S, Restuadi R, Levien R, Girling JE, Holdsworth-Carson SJ, Healey M, Zhu Z, Qi \nT, Wu Y, Lukowski SW et al (2019) Genetic regulation of methylation in human \nendometrium and blood and gene targets for reproductive diseases. Clinical Epigenetics 11: \n49 \nNatanzon Y, Goode EL, Cunningham JM (2018) Epigenetics in ovarian cancer. Semin \nCancer Biol 51: 160-169 \nPearce CL, Templeman C, Rossing MA, Lee A, Near AM, Webb PM, Nagle CM, Doherty \nJA, Cushing-Haugen KL, Wicklund KG et al (2012) Association between endometriosis and \nrisk of histological subtypes of ovarian cancer: a pooled analysis of case–control studies. The \nLancet Oncology 13: 385-394 \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nPharoah PDP, Tsai Y-Y, Ramus SJ, Phelan CM, Goode EL, Lawrenson K, Buckley M, \nFridley BL, Tyrer JP, Shen H et al (2013) GWAS meta-analysis and replication identifies \nthree new susceptibility loci for ovarian cancer. Nature genetics 45: 362-370e3702 \nPhelan CM, Kuchenbaecker KB, Tyrer JP, Kar SP, Lawrenson K, Winham SJ, Dennis J, Pirie \nA, Riggan MJ, Chornokur G et al (2017) Identification of 12 new susceptibility loci for \ndifferent histotypes of epithelial ovarian cancer. Nature Genetics 49: 680 \nPickrell JK, Berisa T, Liu JZ, Ségurel L, Tung JY, Hinds DA (2016) Detection and \ninterpretation of shared genetic influences on 42 human traits. Nature Genetics 48: 709-717 \nPowell JE, Fung JN, Shakhbazov K, Sapkota Y, Cloonan N, Hemani G, Hillman KM, \nKaufmann S, Luong HT, Bowdler L et al (2016) Endometriosis risk alleles at 1p36.12 act \nthrough inverse regulation of CDC42 and LINC00339. Human molecular genetics 25: 5046-\n5058 \nRahmioglu N, Banasik K, Christofidou P, Danning R, Galarneau G, Giri A, MacGregor S, \nMortlock S, Sapkota Y, Schork AJ et al (2018) Large-scale genome-wide association meta-\nanalysis of endometriosis reveals 13 novel loci and genetically-associated comorbidity with \nother pain conditions. bioRxiv DOI:10.1101/406967 \nRoss-Adams H, Ball S, Lawrenson K, Halim S, Russell R, Wells C, Strand SH, Ørntoft TF, \nLarson M, Armasu S et al (2016) HNF1B variants associate with promoter methylation and \nregulate gene networks activated in prostate and ovarian cancer. Oncotarget 7: 74734-74746 \nRossing MA, Cushing-Haugen KL, Wicklund KG, Doherty JA, Weiss NS (2008) Risk of \nepithelial ovarian cancer in relation to benign ovarian conditions and ovarian surgery. Cancer \nCauses & Control 19: 1357 \nRowlands IJ, Abbott JA, Montgomery GW, Hockey R, Rogers P, Mishra GD (2021) \nPrevalence and incidence of endometriosis in Australian women: a data linkage cohort study. \nBJOG: An International Journal of Obstetrics & Gynaecology 128: 657-665 \nRuth KS, Campbell PJ, Chew S, Lim EM, Hadlow N, Stuckey BGA, Brown SJ, Feenstra B, \nJoseph J, Surdulescu GL et al (2016) Genome-wide association study with 1000 genomes \nimputation identifies signals for nine sex hormone-related phenotypes. Eur J Hum Genet 24: \n284-290 \nSaavalainen L, Lassus H, But A, Tiitinen A, Härkki P, Gissler M, Pukkala E, Heikinheimo O \n(2018) Risk of Gynecologic Cancer According to the Type of Endometriosis. Obstet Gynecol \n131: 1095-1102 \nSapkota Y, Steinthorsdottir V, Morris AP, Fassbender A, Rahmioglu N, De Vivo I, Buring \nJE, Zhang F, Edwards TL, Jones S et al (2017) Meta-analysis identifies five novel loci \nassociated with endometriosis highlighting key genes involved in hormone metabolism. \nNature Communications 8: 15539 \nStamp JP, Gilks CB, Wesseling M, Eshragh S, Ceballos K, Anglesio MS, Kwon JS, Tone A, \nHuntsman DG, Carey MS (2016) BAF250a Expression in Atypical Endometriosis and \nEndometriosis-Associated Ovarian Cancer. International Journal of Gynecologic Cancer 26: \n825 \nSugino N, Kashida S, Karube-Harada A, Takiguchi S, Kato H (2002) Expression of vascular \nendothelial growth factor (VEGF) and its receptors in human endometrium throughout the \nmenstrual cycle and in early pregnancy. Reproduction 123: 379-387 \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nSuzuki E, Kajita S, Takahashi H, Matsumoto T, Tsuruta T, Saegusa M (2015) Transcriptional \nupregulation of HNF-1β by NF-κB in ovarian clear cell carcinoma modulates susceptibility to \napoptosis through alteration in bcl-2 expression. Laboratory Investigation 95: 962-972 \nTreloar S, Hadfield R, Montgomery G, Lambert A, Wicks J, Barlow DH, O’Connor DT, \nKennedy S (2002) The international endogene study: a collection of families for genetic \nresearch in endometriosis. Fertility and Sterility 78: 679-685 \nTrevisan CM, de Oliveira R, Christofolini DM, Barbosa CP, Bianco B (2019) Effects of a \nPolymorphism in the Promoter Region of the Follicle-Stimulating Hormone Subunit Beta \n(FSHB) Gene on Female Reproductive Outcomes. Genet Test Mol Biomarkers 23: 39-44 \nTrochet H, Pirinen M, Band G, Jostins L, McVean G, Spencer CCA (2019) Bayesian meta-\nanalysis across genome-wide association studies of diverse phenotypes. Genetic \nEpidemiology 43: 532-547 \nVercellini P, Parazzini F, Bolis G, Carinelli S, Dindelli M, Vendola N, Luchini L, Crosignani \nPG (1993) Endometriosis and ovarian cancer. American Journal of Obstetrics & Gynecology \n169: 181-182 \nVõsa U, Claringbould A, Westra H-J, Bonder MJ, Deelen P, Zeng B, Kirsten H, Saha A, \nKreuzhuber R, Kasela S et al (2018) Unraveling the polygenic architecture of complex traits \nusing blood eQTL meta-analysis. bioRxiv DOI:10.1101/447367 \nWatanabe K, Taskesen E, van Bochoven A, Posthuma D (2017) Functional mapping and \nannotation of genetic associations with FUMA. Nature Communications 8: 1826 \nWei J-J, William J, Bulun S (2011) Endometriosis and Ovarian Cancer: A Review of \nClinical, Pathologic, and Molecular Aspects. International Journal of Gynecological \nPathology 30: 553-568 \nWiegand KC, Shah SP, Al-Agha OM, Zhao Y, Tse K, Zeng T, Senz J, McConechy MK, \nAnglesio MS, Kalloger SE et al (2010a) ARID1A Mutations in Endometriosis-Associated \nOvarian Carcinomas. New England Journal of Medicine 363: 1532-1543 \nWiegand KC, Shah SP, Al-Agha OM, Zhao Y, Tse K, Zeng T, Senz J, McConechy MK, \nAnglesio MS, Kalloger SE et al (2010b) ARID1A mutations in endometriosis-associated \novarian carcinomas. The New England journal of medicine 363: 1532-1543 \nYamaguchi K, Amano Y, Matsumura N, Mandai M, Abiko K, Hamanishi J, Yoshioka Y, \nBaba T, Konishi I (2014) <em>HNF1B</em> contributes to resistance to oxidative stress \nthrough modification of metabolism in ovarian clear cell carcinoma. Gynecologic oncology \n133: 87 \nYarmolinsky J, Relton CL, Lophatananon A, Muir K, Menon U, Gentry-Maharaj A, Walther \nA, Zheng J, Fasching P, Zheng W et al (2019) Appraising the role of previously reported risk \nfactors in epithelial ovarian cancer risk: A Mendelian randomization analysis. PLoS Med 16: \ne1002893 \nYavorska OO, Burgess S (2017) MendelianRandomization: an R package for performing \nMendelian randomization analyses using summarized data. Int J Epidemiol 46: 1734-1739 \nZafrakas M, Grimbizis G, Timologou A, Tarlatzis BC (2014) Endometriosis and ovarian \ncancer risk: a systematic review of epidemiological studies. Front Surg 1: 14-14 \nZhu Z, Zhang F, Hu H, Bakshi A, Robinson MR, Powell JE, Montgomery GW, Goddard ME, \nWray NR, Visscher PM et al (2016) Integration of summary data from GWAS and eQTL \nstudies predicts complex trait gene targets. Nat Genet 48: 481-487 \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\n \nFigures \n \n \n \nFigure 1. Genomic regions associated with risk of endometriosis and epithelial ovarian \ncancer (EOC) histotypes. Regions marked as orange have a posterior probability of model 3 \n(PPA_3) > 0.5 evidence that the same underlying candidate causal variant is associated with \nendometriosis and the ovarian cancer histotype in that region. Regions marked as yellow \nhave  a posterior probability of model 4 (PPA_4) > 0.5 evidence that the region contains a \nvariant associated with endometriosis and an independent variant associated with ovarian \ncancer histotype. All regions shaded in orange contained genome-wide significant loci in the \ncross-trait meta-analysis except Chr 3:126215130-128194265. \n \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\n \nFigure 2. Functional annotation of SNPs associated with risk of endometriosis and ovarian \ncancer. (a) Histogram of number of non-redundant SNPs for all histological subtypes that \noverlap n biofeatures. Inset panel, histogram of number of non-redundant SNPs for all \nhistological subtypes that overlap ≥ 1 biofeature. (b) Proportion of loci shared between \nendometriosis and each EOC histotype containing SNPs that overlap at least one biofeature. \n(c) A promoter SNP at the VEZT/FGD6 locus overlaps 10 biofeatures and intersects with an \nactive open region of chromatin that lies in a bidirectional promoter associated with these two \ngenes. Biofeatures are shown as peaks on the ATAC-seq, H3K27ac and ChIP-seq tracks for \nprimary tissues (PT) and cell lines (CL). Endo, endometriosis. Specimens are primary tissues \nunless otherwise indicated. The grey shaded area highlights peaks overlapping rs6538618. \nBoxplots show the association between rs6538618 genotypes and expression of VEZT and \nFGD6 in endometrium. \n \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nTables \nTable 1. Genetic correlation ( rg) between endometriosis and epithelial ovarian cancer (EOC) \nhistotypes. \nEOC Histotype rg (se) P-value \nClear cell 0.71 (0.26) 0.007 \nEndometrioid 0.48 (0.20) 0.016 \nHigh-grade Serous 0.19 (0.09) 0.033 \nLow-grade Serous NA NA \nLow malignant potential serous 0.88 (0.85) 0.401 \nMucinous -0.18 (0.15) 0.227 \nse = standard error \n \nTable 2. Mendelian randomization results considering genetic liability to endometriosis as the \nexposure and epithelial ovarian cancer (EOC) histotypes as the outcome. \nEOC Histotype MR Method OR (95% CI) P-value \nHigh Grade Serous IVW 1.22 (1.07-1.38) 0.002 \n \nWeighted \nMedian 1.16 (1.02-1.32) 0.025 \n MR-Egger 1.23 (0.82-1.86) 0.319 \nLow Grade Serous IVW 1.27 (0.96-1.67) 0.091 \n \nWeighted \nMedian 1.27 (0.87-1.84) 0.212 \n MR-Egger 0.86 (0.35-2.09) 0.742 \nLow Malignant Potential \nSerous IVW 1.45 (1.17-1.79) 0.001 \n \nWeighted \nMedian 1.52 (1.16-1.99) 0.003 \n MR-Egger 1.42 (0.71-2.83) 0.323 \nMucinous IVW 1.24 (1-1.53) 0.046 \n \nWeighted \nMedian 1.03 (0.77-1.39) 0.821 \n MR-Egger 0.83 (0.42-1.63) 0.583 \nEndometrioid IVW 1.66 (1.42-1.93) 1.4E-10 \n \nWeighted \nMedian 1.58 (1.27-1.97) 3.0E-05 \n MR-Egger 1.63 (1-2.67) 0.051 \nClear Cell IVW 2.59 (2.09-3.21) 2.8E-18 \n \nWeighted \nMedian 2.48 (1.82-3.39) 9.6E-08 \n MR-Egger 2.54 (1.28-5.02) 0.007 \nOR = Odds Ratio    \n95% CI = 95% Confidence Interval   \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nTable 3. Number of significant SNPs and genomic loci identified in the epithelial ovarian \ncancer (EOC) histotype and endometriosis cross-trait meta-analyses. \nEOC Histotype that \nwas combined with \nEndometriosis \nNo. SNPs with \nABF > 4 in \ncombined \ndata \nNo. SNPs with \nRE2C  \nP-value < 5x10-8 in \ncombined data \nNo. of genomic loci \nsignificant in both \nMetABF and RE2C* \nClear cell 1055 336 14 \nEndometrioid 759 309 6 \nHigh-grade serous 2814 1851 13 \nLow-grade serous 733 237 3 \nLMP serous 1114 514 5 \nMucinous 899 365 5 \n* The number of genomic loci significant in both  MetABF and RE2C reported here is the \nset of loci where the lead SNPs achieved ABF > 4 and RE2C P -value < 5 x 10 -8 in the \ncombined data and had nominal evidence of association (P < 0.05) in each individual GWAS \ndata set (i.e., in the endometriosis data set and in the corresponding EOC histotype data set). \nIndependent loci were identified by linkage disequilibrium-based pruning at r2 < 0.6. \n \nTable 4. Lead SNPs in genomic loci that demonstrated shared associations with an epithelial \novarian cancer (EOC) histotype and endometriosis from the RE2C meta-analysis. \nrsID Chr Posa ABF Be P-value RE2Cp Pval_Ovarian Pval_Endo \nNearest gene or \ngene with \nfunctional \nevidence \nClear cell ovarian cancer + Endometriosis \nrs61768001 1 22465820 10.92 0.13 9.54E-14 1.00E+00 2.03E-02 1.59E-12 LINC00339 \nrs11674184 2 11721535 13.27 -0.11 3.86E-16 1.00E+00 3.02E-03 3.19E-14 GREB1 \nrs10167914 2 113563361 8.07 0.11 9.57E-11 1.00E+00 2.59E-04 4.94E-08 IL1A \nrs4516787 4 56010165 10.97 -0.11 9.10E-14 1.00E+00 8.20E-04 1.88E-11 KDR \nrs1311245 5 64272107 5.5 0.07 4.55E-08 1.00E+00 2.27E-02 5.58E-07 CWC27 \nrs1971256 6 151816011 6.17 0.09 9.57E-09 1.00E+00 3.20E-02 9.68E-08 CCDC170 \nrs17803970 6 152553718 6.83 -0.15 1.62E-09 1.00E+00 0.002401 9.82E-08 SYNE1 \nrs71575922 6 152554014 8.04 0.12 1.01E-10 1.00E+00 4.34E-04 2.02E-08 SYNE1 \nrs12700667 7 25901639 7.27 0.09 6.58E-10 1.00E+00 1.18E-02 1.51E-08 AK057379 \nrs78103255 8 75311331 6.08 -0.09 1.19E-08 1.00E+00 6.64E-04 2.47E-06 GDAP1 \nrs566679*d 9 22634893 6.26 0.07 2.60E-05 2.10E-08 8.64E-03 6.62E-08 LINC01239 \nrs7309252 12 95687497 5.69 0.07 2.84E-08 1.00E+00 4.53E-03 1.07E-06 VEZT \nrs11651755* 17 36099840 6.08 0.05 4.52E-05 1.00E+00 6.78E-09 2.02E-02 HNF1B \nrs8069263 17 46286778 5.54 0.07 4.10E-08 8.03E-08 3.86E-02 3.88E-07 SKAP1 \nEndometrioid ovarian cancer + Endometriosis \nrs56318008 1 22470407 10.11 0.12 6.77E-13 1.00E+00 3.37E-02 3.50E-12 LINC00339 \nrs495590 1 172122809 5.7 0.08 2.97E-08 1.00E+00 2.00E-04 2.42E-05 DNM3 \nrs1971256 6 151816011 7.36 0.1 5.28E-10 1.00E+00 1.26E-03 9.68E-08 CCDC170 \nrs6475610 9 22141894 8.39 0.08 4.33E-11 1.00E+00 7.29E-03 1.73E-09 CDKN2B-AS1 \nrs11031005 11 30226356 7.39 -0.11 4.94E-10 1.00E+00 1.06E-03 1.03E-07 FSHB \nrs10445377 17 46214168 6.66 0.08 2.71E-09 1.00E+00 1.95E-03 3.20E-07 SKAP1 \nHigh-grade serous ovarian cancer + Endometriosis \nrs12037376* 1 22462111 10.99 0.1 3.15E-13 3.06E-14 3.51E-03 1.04E-12 LINC00339 \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\nrs7570979 2 11717429 7.76 0.08 1.91E-10 1.00E+00 0.006594 1.43E-09 GREB1 \nrs13000026 2 165558884 7.03 -0.07 1.08E-09 1.00E+00 1.10E-05 2.27E-05 COBLL1 \nrs1250244* 2 216297796 5.57 -0.07 1.36E-07 1.61E-08 2.69E-02 8.73E-08 FN1 \nrs6908034*d 6 19773930 7.23 0.04 0.002250956 1.65E-09 2.21E-02 2.09E-09 ID4 \nrs111610638 6 152449994 5.54 -0.15 3.65E-08 1.00E+00 1.84E-03 3.90E-06 SYNE1 \nrs1981046 9 22173407 5.44 -0.06 4.66E-08 1.00E+00 2.31E-02 1.03E-07 CDKN2B-AS1 \nrs635634* 9 136155000 10.64 0.09 9.40E-13 6.92E-14 2.25E-11 3.39E-04 ABO \nrs7084454 10 21821274 10.39 0.08 3.50E-13 1.00E+00 2.77E-09 9.06E-06 MLLT10 \nrs11658063*d 17 36103872 7.87 -0.03 0.014979392 2.98E-10 3.63E-10 1.83E-02 HNF1B \nrs62065444* 17 43565599 11.11 0.12 2.08E-13 2.68E-14 1.20E-13 4.77E-02 PLEKHM1 \nrs7217120* 17 46484755 13.5 0.09 7.96E-15 1.08E-16 2.08E-14 3.69E-04 SKAP1 \nrs10048393 18 3476253 5.62 0.06 3.09E-08 1.00E+00 2.12E-05 3.23E-04 AX721193 \nLow-grade serous ovarian cancer + Endometriosis \nrs77294520 2 11660955 10.72 0.15 1.39E-13 1.00E+00 0.04698 9.91E-13 GREB1 \nrs584336*d 6 152616173 6.35 0.07 4.28E-06 1.61E-08 1.11E-02 4.17E-08 SYNE1 \nrs10445377 17 46214168 5.72 0.08 2.69E-08 1.00E+00 1.90E-02 3.20E-07 SKAP1 \nLMP serous ovarian cancer + Endometriosis \nrs4654785 1 22491843 6.78 0.09 2.18E-09 1.00E+00 0.03783 1.89E-08 LOC105376850 \nrs10748858* 10 105639514 5.81 0.07 6.06E-07 1.30E-08 1.09E-06 5.31E-04 OBFC1 \nrs11031005 11 30226356 6.24 -0.11 8.14E-09 1.00E+00 2.73E-02 1.03E-07 FSHB \nrs10445377 17 46214168 7.46 0.08 4.06E-10 1.00E+00 4.56E-05 3.20E-07 SKAP1 \nrs35713035* 17 46501710  6.43 0.09 1.75E-08 1.74E-09 4.32E-06 3.60E-05 SKAP1 \nMucinous ovarian cancer + Endometriosis \nrs11674184 2 11721535 12.11 -0.1 6.13E-15 1.00E+00 3.23E-02 3.19E-14 GREB1 \nrs6546324 2 67856490 5.91 -0.08 1.71E-08 1.00E+00 1.76E-02 3.02E-07 LINC01812 \nrs10167914 2 113563361 6.98 0.1 1.36E-09 1.00E+00 7.23E-03 4.94E-08 IL1A \nrs4849174* 2 113973467 12.14 0.09 2.66E-09 2.00E-15 2.53E-14 2.89E-03 PAX8 \nrs67808862* 3 138849543 10.26 0.07 5.94E-07 2.53E-13 2.14E-13 4.64E-02 BPESC1 \nABF: log10 approximate Bayes factor; LMP, low malignant potential \nBe: Estimated beta coefficient from the fixed effects model. \nP-value: Fixed effect model P-value \nRE2CP: RE2C P-value (RE2Cp is 1 for SNPs where there is little or no evidence of heterogeneity across the two traits and for such \nSNPs the fixed effects model and its corresponding association P-value becomes the model of choice) \n*SNP with significant heterogeneity and results presented from independent MetABF and RE2C random effects model. \na: Build 37 positions. \nd: Direction of effect is different for each trait  \n \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\n 40 \nTable 5. GWAS-pw results for analyses between epithelial ovarian cancer (EOC) histotypes \nand endometriosis. Posterior probabilities of GWAS-pw models. \nEOC Histotype Chr Region (posa) PPA_1 PPA_2 PPA_3 PPA_4 \nClear Cell 1 21736898:23086667 0.3 0 0.69 0.01 \nClear Cell 2 10298766:12418752 0.07 0 0.93 0 \nClear Cell 2 110857126:113921639 0.01 0 0.99 0 \nClear Cell 4 55429886:56547412 0.03 0 0.97 0 \nClear Cell 5 63968304:65910972 0.26 0 0.72 0.01 \nClear Cell 6 150256048:151912653 0.31 0 0.63 0.01 \nClear Cell 6 151912703:153093958 0.06 0 0.93 0 \nClear Cell 7 25077628:25909208 0.32 0 0.67 0.01 \nClear Cell 8 73817199:75444858 0.04 0 0.95 0 \nClear Cell 9 22206559:24157796 0.17 0 0.81 0 \nClear Cell 12 94514787:96019818 0.13 0 0.85 0 \nClear Cell 17 34812273:36808793 0 0.01 0.97 0.01 \nClear Cell 17 45876022:47516523 0.37 0 0.61 0.01 \nEndometrioid 6 150256048:151912653 0.42 0 0.54 0 \nEndometrioid 17 45876022:47516523 0.46 0 0.53 0.01 \nHigh Grade \nSerous 1 21736898:23086667 0.16 0 0.7 0.15 \nHigh Grade \nSerous 2 10298766:12418752 0.37 0 0.06 0.57 \nHigh Grade \nSerous 2 165178853:167160029 0.02 0 0.89 0.03 \nHigh Grade \nSerous 6 151912703:153093958 0.35 0 0.05 0.6 \nHigh Grade \nSerous 9 135298917:137040737 0 0 0.97 0.03 \nHigh Grade \nSerous 10 19717815:22772115 0 0 0.99 0.01 \nHigh Grade \nSerous 17 34812273:36808793 0 0.02 0.43 0.55 \nHigh Grade \nSerous 17 43056905:45875506 0 0.04 0.18 0.78 \nHigh Grade \nSerous 17 45876022:47516523 0 0 0.06 0.94 \nHigh Grade \nSerous 18 1943138:3890554 0.02 0 0.76 0.06 \nLMP Serous 10 104380686:106694980 0.01 0.01 0.92 0.04 \nLMP Serous 17 45876022:47516523 0.02 0 0.73 0.25 \nMucinous 2 113922276:116772246 0 0.11 0.87 0.03 \nMucinous 3 126215130:128194265 0.4 0 0.53 0.02 \nPPA_1: posterior probability of model 1 [association only to Endometriosis] \nPPA_2: posterior probability of model 2 [association only to EOC] \nPPA_3: posterior probability of model 3 [shared association to both phenotypes] \nPPA_4: posterior probability of model 4 [two distinct associations, one to each phenotype] \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\n 41 \na: Build 37 positions. \n \nTable 6. Number of SNPs overlapping zero or more than zero consensus peak sets. Related \nto Supplementary Figure 4. \nComparison nOverlap = 0 nOverlap > 0 Total number of \nSNPs \nEndo_CCOC 208 41 (16.5%) 249 \nEndo_EnOC 109 7 (6.0%) 116 \nEndo_HGSOC 388 56 (12.6%) 444 \nEndo_LGSOC 29 1 (3.3%) 30 \nEndo_LMPSOC 191 7 (3.5%) 198 \nEndo_MOC 74 33 (30.8%) 107 \n \n \n \nTable 7. Summary of evidence from the bivariate meta -analysis, GWAS -pw, overlap of \nbiofeatures, fastBAT and Summary -data-based Mendelian Randomisation (SMR) for loci \nassociated with both endometriosis and epithelial ovarian cancer (EOC) histotypes. Cells \nshaded or ange contain EOC histotypes with a posterior probability of a shared variant \nassociation with endometriosis of >0.5, those shaded yellow contain EOC histotypes with a \nposterior probability of two distinct associations, one to each phenotype, of >0.5. The b lue \ncolour gradient in the feature overlap column corresponds to the maximum number of feature \noverlaps identified for a variant in that locus. \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint \n\n 42 \n \n . CC-BY-ND 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted July 2, 2021. ; https://doi.org/10.1101/2021.06.28.21259290doi: medRxiv preprint","source_license":"CC0","license_restricted":false}