Enhancing Genetic Association Power in Endometriosis through Unsupervised Clustering of Clinical Subtypes Identified from Electronic Health Records

23

Endometriosis af fects 10% of repr oductive-age women, and yet, it goes undiagnosed 24 for 3.6 years on average after sympto ms onset. Despite large GWAS meta-a nalyses (N > 750 ,000), 25 only a few d ozen causal loci have been identified. We hypothesized th at the cha llenges in 26 identifying causal genes for endometriosis stem fr om heterogeneity across clinical and biological 27 factors underlying end ometriosis diag nosis. 28

We extracted known endometriosis risk factors, symptoms, and co ncomitant conditions 29 from the Penn Medicine Biobank (PMBB) and perfo rmed unsupervised spectr al clust ering on 4,078 30 women with endometriosis. The 5 clusters were characterized by utilizin g additional electronic 31 healt h record (EHR ) variables, such as endometriosis-related comor bidities a nd confi rmed surgical 32 phenotypes. From four EHR-linked ge netic da tasets, PMBB, e MERGE, AOU , a nd UKBB, we extracted 33 lead variants and tag variants 39 known end ometriosis loci f or as sociati on testing. We meta-34 analyzed ancestry-stratified case/cont rol tests for each locus and cluster in addition to a positive 35 control (T otal N endometriosis cases = 10,108) . 36

We have designated the five subtype clust ers as pain comorbidities, uterine disorders, 37 pregnancy complications, c ardiomet abolic comorbidities, and EHR-asy mptomatic based on 38 enriched features from each group. One locus, RNLS, surpassed the gen ome-wide significan t 39 threshold in the positive control. Thirteen more loci reached a Bon ferr oni t hreshold of 1.3 x 10 -3 40 (0.05 / 39) in the positive control. The cluster-stratified tests yield ed more si gnificant associations 41 than t he positive control for anywhere from 5 to 15 loci depending on t he cl uster. Bonferroni 42 significant loci were identified for f o ur out of five clusters, including WN T4 and GREB1 f o r t h e 43 uterine disorders cluster , RNLS for the cardiometabolic c luster, FSHB for the pregnancy 44 complications cluster, and SYNE1 a n d CDKN2B -AS 1 f or the EHR-asymptoma tic cluster. This study 45 enhances our understanding of the clinical presentation pat terns of end ometriosis su btypes, 46 showcasing the inn ovative approach employed to investigate this complex disease. 47 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 3 Abbreviations 48 AOU All of Us Bio bank 49 eMERGE electronic medical record and genomic s network 50 EHR electronic healt h record 51 GWAS genome-wide ass ociation study 52 ICD international classification of diseases 53 PMBB Penn Medicine biobank 54 UKBB United Kingdom bi obank 55

56 Endometriosis, a complex gynecological condition affects 10% of women of reproductive 57 age globally and more than 50% of wo men with infertility (1), ye t it often go es either undiagnosed 58 or misdiagno sed, leading to delay ed diagnoses and delivery of eff ective th erapy (2,3). 59 Endometriosis is primarily characteriz ed by the presence of endometrial-like tissue outside of the 60 uterus. For managing the condition without surgery, th e main trea tments i nclude pain reli ef and 61 hormone-based therapies, neither of which are curative. A notable nu mber o f w omen with 62 endometriosis receive opioids for pai n management , despite th e need for more sustainable and 63 effective treatm ent options (4,5). On the other hand, hormonal therapies m ay have limi tations to 64 utiliza tion due to severe side effects o r a desire to become pregnant. Typically , the trea tment for 65 endometriosis often includes b oth medical and surgical approaches, however 30-50% o f patients 66 with severe endome triosis may require a second surgery within 3-5 years (6). The most 67 comprehensive surgical management i nvolves a hysterectomy with bilatera l salpingoophorectomy 68 (7).Treatment and hea lth care visits accumul ate many direc t and indirec t costs for w omen with 69 endometriosis. The estimated economic cost o f endometriosi s in the US is ~$ 10k per patient which 70 is ~14% higher than that of diabetes ( 8), and does not include the costs patients incur by having t o 71 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 4 miss wo rk days because o f their symptoms. In total, endome triosis pres ents a high economic 72 burden that exceeds $22 billion in th e U.S. a lone (9). The condition not o nly imposes significant 73 costs but also involves severe sym ptoms, del ayed di agnosis, limited tr eatment options, and 74 financial strain: challenges tha t could be significantly mitigated with a more detaile d 75 understanding of the disease. 76 Electronic Healt h Records (EHRs) represent a rich, yet underuti lized , da ta source for 77 capturing the phenotypic spec tru m of endometriosis (10). Although the symptoms for 78 endometriosis can be quite severe , inc luding chronic debilita ting pain, dyspa reunia, and inferti lity , 79 the average ti me to diagnosis is 4.5 years (11), in part because the only way to definitively diagnose 80 endometriosis is by surgical observati on o f endometrial lesions gro wing out side of the uterus (e.g. 81 abdominal cavity , pel vis, ovaries, e tc.) (12). The variabilit y in symptoms an d disease presentation 82 adds to the difficulty of diagnosis and h inders the optima l use of ele ctronic he alth records (EHRs) in 83 research for accurately identifying affected individua ls and control subjects (13–15), which is 84 critical for understanding the disease a nd advancing treatmen t strategies. T he depth and breadt h of 85 EHR data provide a unique opport unity to apply unsupervised learnin g techniques fo r the 86 identification of distinct phenotypic clusters that ma y correspond to clinical subtypes of 87 endometriosis. Such an approach a li gns with precision medicine's goal t o tailor diagnosis and 88 treatment strategies to individu al pati ent charac teristics, pot ential ly re veali ng novel insights into 89 the disease's pathophysiology. 90 Better understanding of the disease m echanisms of endometrio sis c ould lea d to improved 91 diagnostic practices, reducing costs to the heal thcare system and improving quality of life through 92 treatment and earl ier diagnosis fo r patients. In spite of the preval e nce and severity of 93 endometriosis, etiology of endometr iosis is still poorly understood. The pursuit thus far of 94 biomarkers and drug targets base d on genetic contributions of dise ase in patients with 95 endometriosis has mainly include d g enome-wide association studies to id entify genetic variants 96 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 5 contributing to the disease (16 ,17). Twin studies have estimat ed the heritabili ty of endometriosis to 97 be 47.5% (18), and common variants are estima ted to contributed 26% of phenotypic variance 98 (19), but the largest GWAS to-date ( N > 750,00 0, 60 ,674 cases) has only explained 9% of the 99 phenotypic variance (17). Although these recent advances in genomic st udies have promised 100 insights into the underlying genetic mechanisms of endometriosi s, yet th e heterogeneity of the 101 disease presentation has consistently complicated these efforts. Tradition al genetic association 102 studies have struggled to untangle th e intricate web of genotypic and pheno typic diversity within 103 endometriosis patients, leading to a critical need for innovative approaches to dissect the disease's 104 complexity. 105 We hypothesized that underlying clinical heterogeneity is obscuring the genetic 106 mechanisms and preventing large-scale genetic studies from explaining mo re of the heritability . 107 Endometriosis causes a wide range of symptoms and conc omitant conditions, including severe 108 chronic pain, gastrointestinal inflam mation, and infertilit y. Addi tionally , many symptoms of 109 endometriosis are shared between ot her gynecological diseases su ch as pr imary dysmenorrhea, 110 ovarian cysts, and pelvic infla mmato ry disease; making symptom-based diagnosis challenging 111 (20,21). Recent studies hav e highlight e d the importance of complex disease s ubtyping in improving 112 our understanding of the genetic me chanisms underlying endome triosis. For example, a recent 113 study on polycystic ovary syndrom e (PCOS) used unsupervised clust ering to identify three 114 subtypes of PCOS based on lab and biometric valu es before conducting genome-wide associatio n 115 study for each subt ype (22). This ap proach allowed for a more nu anced understanding of the 116 genetic basis o f PCOS and could be a pplied to endometriosis to identify s ubtypes with distinct 117 genetic mechanisms. Building on the premise that a more nuanced understanding of endometriosi s 118 subtypes could unlock new genet ic a ssociations, our study leverages unsu pervised, phenotypic 119 clustering analysis of EHR data to s ystematica lly iden tify and characteriz e clinical subtypes o f 120 endometriosis. By dissecting the he te rogeneity inherent in th e disease, we aim to increase the 121 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 6 power o f genetic ass ociation analyses, facilitating the iden tification of su btype-specific disease 122 mechanisms. This approach not only promises to enhance our understan ding of end ometriosis 123 genetics but also to refine diagnostic criteria and inform more targeted a nd effective treat ment 124 strategies. 125 In conclusion, the complex nature of end ometriosis, with its diverse s ymptoms and 126 overlapping features with other gynecological diseases, presents challenges for under standing its 127 genetic mec hanisms. In this manuscrip t, we detai l the methodology and findings of our study, which 128 integrates unsupervised phenotypic c lustering with subsequent genetic association analyses fo r 129 each iden tified endometriosis subtyp e. By doing so, we aim to bridge the gap between c linical 130 observations and genetic research in endometriosis, providing a roadmap for future studies to 131 explore the gene tic und erpinnings of this complex disease wit h renewed clar ity and precision. This 132 deeper understanding ma y pave the way for more targeted and personalized ap proaches to 133 diagnosis, treatment, and manage men t of this debilitating condition. Furthe r research and large-134 scale genetic studies are needed to fully elucidat e the genetic archite cture of endometriosis and its 135 subtypes, ultima tel y leading to improv ed outcomes f or a ffected individuals. 136

137 Datasets Used for Sub-phenotyping and Genetic Association 138 The Penn Medicine Biobank (PMB B) is the University of Pennsylvania’s healt h system-139 based bio bank which consists of a bout 250,000 consented par ticipants, with 4 3,624 of those having 140 imputed genotype da ta (imputed to TOPMED reference panel) linked with t heir electronic heal th 141 record (EHR) history. The P MBB is an electronic he alt h record (EHR)-linked biobank that integrates 142 a wide variety of health-related information, including diagnosis codes, lab oratory measurements, 143 imaging data, and lifestyle information, with genomic and bi omarker data. T he PMBB is one o f the 144 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 7 most diverse medical biobanks, with approximately 30% of participants being of non -European 145 ancestry. This diversity is crucia l for ensuring that research findings are applicable to a broad range 146 of populations. The biobank also bene fits from a median of seven years of lo ngitudinal da ta in t he 147 EHR, providing valuable information on participants' health histories (2 2). For our study, we 148 treated th e PMBB as two distinct datasets: those without and those with ge notype data. EHR data 149 from the non-genotyped PMBB were used for cluster derivat ion whereas the genotype d PMBB 150 cohort was used in the genetic analyses. 151 The Electronic Medical Re cords and Genomics (eMERGE ) network i s a National Human 152 Genome Research Insti tute-funded consortium engaged in the deve lopment of methods and best 153 practices for using the electronic me dical record as a tool for genomic re search. The e MERGE 154 network is a publicly-availabl e datase t with contributions fr om multiple hea l th systems within the 155 United Stat es which contains about 100,000 partici pants with linked h ealt h records and imputed 156 genomic data (imputed to HRC reference panel) (23). The eMERGE consortium validat ed th e 157 hypothesis that clinica l data derived from electronic medical records can b e used successfully for 158 complex genomic analysis of disease susceptibility across diverse patient populations (24). The 159 eMERGE network has shown the effici ency that can resul t from the use of el ectronic health record 160 data. 161 The All of Us (AOU) Researc h Program is an ini tiativ e crea ted by the N IH to recruit 162 demographically diverse indi viduals to the largest US-based biobank to-date. Recruitment began in 163 2018, and since th en, over 400, 000 pe ople have signed up and submitted ba seline questions (25). 164 245,388 of them have short-read whol e genome sequence dat a, col lecti vely representing over one 165 billion genetic variants (26). Participa nts’ EHRs are contributed to the AOU data processing center 166 using the Sync for Science platform (27 ), which works with EHR vendors such as Epic and Cerner to 167 collate structured pati ent data for rese arch use (28). 168 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 8 The UK Bio bank (UKBB ) is a large and comprehensive dataset tha t provides v aluable 169 resources for researchers studying a w ide range o f health-related topics. The UKBB is a p opulation-170 based publicly available datase t consis ting of about 500,000 UK citizens with EHR data, heal th 171 survey data, and imputed genotypes. T he UK Bio bank has per fo rmed genome- wide genotyping on 172 all participants using the UK Bio bank A xiom A rray (29). This array directly measures 173 approximately 850,000 varian ts, and more than 90 million variants are imp uted using the 174 Haplotype Reference Cons ortium and UK10K + 1000 Genomes reference pan els. 175 All four of the biobanks mentioned above (PMBB, eMERGE, AOU, and UK BB) utilize th e 176 Observational Medica l Out comes Par tnership Common Data Model (OMO P-CDM) to represent 177 structured EHR data in a harmonized format (30). For this study, we utiliz ed women with ICD-178 diagnosed endometriosis in the non-genotyped PMBB cohort (N endo = 4,0 78) as t he d erivation 179 dataset for the clinical subtypes. For deeper characterization of our subtypes, we perfo rmed chart-180 reviews on 682 randomly selected end ometriosis cases from the genotyped PMBB. Then, we meta-181 analyzed women from the genotyped PMBB (N = 20 ,697 , N endo = 1, 198), six non-pediatric sites 182 w i t h i n t h e e M E R G E n e t w o r k ( N = 5 1 , 8 0 0 , N endo = 2 ,243), the AOU research p rogram (N = 108 ,098 , 183 N endo = 2 ,126), and UKBB (N = 2 61,8 24, N endo = 4,45 1) to form our main genetic analysis test set (N = 184 442,419 , N endo = 10,01 8). 185 Each of the biobanks projected thei r samples onto th e t housand genomes reference 186 population and performed clustering t o assign genetically inferred ancestry l abels corresponding t o 187 those from the thousand genomes pr oject (31). We restricted our genetic association analyses to 188 the groups which had substanti al sam ple sizes, whic h were those with high similarity t he AFR and 189 EUR thousand genomes superp opulati ons. We will refer t o those groups usi ng AFR and EUR fr om 190 here on out. 191 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 9 Extraction of Endometriosis-Related Clinical Features 192 Patients wit h endometriosis have he t erogeneous clinical presentations; t he re are a wide 193 variety of ass ociated symptoms, risk f actors, and comor bidities. We fir st determined participants’ 194 case-control status o f endometriosi s us ing structured EHR data: ICD-9 and I C D-10 billing codes 617 195 and N80, respectiv ely . Then for endom etriosis cases, we determined whether each individual h ad a 196 history of end ometriosis- related clini cal features. In tota l, we extracted 3 9 ICD-based features 197 (Table S1): 9 ICD -based anatomical subtypes, 14 comorbidities, 8 sympto ms, and 8 pregnancy-198 related p henotypes. We sele cted onl y symptoms, comorbidities, and pregna ncy-related conditions 199 for clustering, removing th e 9 an atomical subtyp es to be used do wnstream in cl uster 200 characterization. We further restrict ed these conditions to those with a prevalence amongst 201 endometriosis cases in the subtype dataset of at least 5%, lea ving us wit h 17 features for the 202 clustering analysis (Figure S1). 203 Unsupervised Clustering 204 We tested four popular methods for unsupervised clus tering: spectral clus tering, density-205 based spatial clustering of applications with noise (D BSCAN), hierarchical agg lomerative clust ering, 206 and k-means clustering. Spectral clust ering identifies clusters by decomp osi ng a dataset’s affinity 207 matrix into its eigenvectors and then clustering in the eigenvector space using QR clustering 208 algorithm (32,33) . DBSCAN is an algorithm which i dentifies dense regions of data points to discover 209 clusters (34). Hi erarchical agglomerat ive cl ustering is an unsupervised c lassification method th at 210 uses a pairwise distance matrix to iterativel y merge nearby points to gether (35). K-means 211 clustering randomly initi alizes centroids for each cluster and then alterna t es between assigning 212 data points to their nearest centroid an d adjusting the centroids until converg ence (36). 213 In addition to choosing an algorithm , a common struggle with unsupervise d clustering is 214 choosing a target number of clusters in a non-arbitrary way. We used several empirical metrics for 215 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 10 this: silhouette score, distortion score, and a metric we developed to represent the “evenness” of 216 clusters. The si lhouet te score is a metr ic which considers both intra- and inter-cluster distances to 217 assess tightness within a c luster and distance between c lusters; high er silh ouette scores indicate 218 better quality clust ers. The distortion score is the sum o f s quared erro r s with respect to the 219 centroid of each clust er, t hus it is desir ed to minimiz e distortion. Our evennes s metric, optimi zed by 220 minimization, was defined as the fractional difference between the size o f th e largest and smallest 221 clusters. We measured th ese metrics across tests for 2 -20 clusters for each of the f our clustering 222

(except f or DBSCAN which au tomatical ly infers the optimal number of clusters). 223 Characterization of Unsupervised Clusters 224 After identifying clinical clusters wit hin our ob servation dataset, our obj ective was to 225 delineat e th eir characteristics . We p erformed two-p opulation z-score pro portion tests (37) to 226 determine if the rat es of input conditions were sig nificantly different on a cluster-vs-other-clusters 227 basis. For our training set (th e non-genotyped PMBB, N endo = 4 ,078), we examined two sets of 228 features for the z-score tests: the 1 7 input features as wel l as ICD-based anatomical subtypes of 229 endometriosis including adenomyosi s, endometrioma, superficial lesions, and deep lesion s 230 (Supplementary Table S2). For c haracterizing our cl usters, we also uti li zed a chart-reviewed 231 dataset of 682 genotyped PMBB patients with endometriosis IC D codes. Th e features considered 232 here were confirmed endometrio sis and adenomyosis status and chart-abstracted symptoms, 233 comorbidities, and surgical phenotyp es (Supplementary Tabl e S3). By co nsidering the cl uster-234 specific dif ferences in these EHR-der ived features among the two datasets, we could observ e 235 patterns in clinical presentation. Based on these patterns, we as signed labels t o each cluster. 236 Cluster-Stratified Candidate Gene Association Testing 237 To identify genetic heterogeneity amo ng the varied clinical presentations of endometriosis, 238 we performed cluster-stratified, ancest ry-stratified candidate gene association studies. Using P LINK 239 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 11 2.0 (38), we extract ed single nucl eotid e polymorphisms (SNPs) in LD (kb dist ance 240 0.1) with 39 au tosomal lea d SNPs rep orted in the most recent endometriosis GWAS(17). LD was 241 computed based on the thousand geno mes reference panel (31). Cluster phenotypes were assigned 242 for PMBB, eMERGE, and AOU using a K-Nearest neighbor s’ classifier (39) wi th K=3 on the same 17 243 ICD-based features. For each stud y, we employed a linear mixed mod el regression method 244 employed in SAIGE (40) to t est for associations between genotypes and case -control status. Cases 245 were females with endometriosis from one cluster and controls were biological femal es with no ICD 246 history of endometrio sis. In the regres sion models we included the first four principal components, 247 age, and batch indicators (eMERGE o nly) as covariates. The ancestry-stratified results of these 248 studies were then met a-analyz ed usin g Plink 1.9 (41) for each of the cluster -phenotypes. We also 249 tested a baseline overall endometriosis (cases from all c lusters combined) a s a positive control to 250 identify how many known loci we wer e able to replicate . Because m ultip le g enetic ancestry groups 251 were included , we c hose a random-effects meta-anal ysis, which is more robust to heterogeneity 252 (42). 253

254 Derivation, Study, and Validation Datasets 255 This study uti lized five datase ts t o investigate th e genetic mechanis ms underlying 256 endometriosis and its su btypes. The datasets used were endometriosis cases in the non-gen otyped 257 PMBB for the deriva tion of clusters, a c hart-reviewed endome triosis cohort to help charac terize the 258 clusters, the genotyped PMBB, six site s within t he e MERGE network, AOU , a nd UKBB for genome-259 wide association analyses (See Met h ods). The sample siz es for each coh ort, the mean age at 260 diagnosis, the number o f cases and c ontrols, and the mean age at the tim e of data pull for each 261 cohort are shown in Ta ble 1. See Me t hods for details on each of the four d atasets. By leveraging 262 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 12 these da tasets, t he st udy aimed to i dentify endometriosis subtypes and gain insights into t he 263 genetic factors as sociated with endom etriosis and its subtypes. 264 265 266 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 1 3 Table 1: Cohort sample size and average age of cases and controls for the datasets used in this 267 analysis. Age was considered the age as of when the EHR data were collected. 268 Datase t Endometriosis N (AFR / EUR) Mean Age (SD) Cluster Derivation Set: Non-Genotyped PMBB Cases 4,078 (NA) 49.9 (13.3) Genetic Association Sets: AOU Cases 2,126 (542 / 1,584) 52.2 (12.8) Controls 108,099 (31,435 / 76,664) 56.8 (16.8) eMERGE Cases 2,243 (353 / 1,890) 59.9 (14.6) Controls 49,557 (9,934 / 39,623) 59.7 (23.4) PMBB Cases 1,198 (562 / 636) 54.2 (12.9) Controls 19,493 (6,524 / 12,969) 60.0 (17.8) UKBB Cases 4,541 (112 / 4,429) 51.5 (7.5) Controls 257,283 (4,524 / 252,759) 56.6 (8.0) Meta-Analysis Totals: META Cases 10,108 (1,569 / 8,539) 53.9 (11.3) Controls 434,432 (52,417 / 382,015) 57.1 (13.6) Derivation of Unsupervised of Clusters 269 Unsupervised clustering was performed in non-genotyped PMBB dataset of 4,078 wome n270 with EHR-diagnosed endometriosis using 17 clinical features (supplementary figure S2). We teste d271 four methods for unsupervised clustering as well as 19 values for the number of clusters (K=2-20 )272 and measured three metrics to empirically choose a clustering method and number of cluster s273 (Figure 1). 274 275 Figure 1: testing various clustering algorithms and K-values to empirically choose an optimal 276 method. The three metrics shown are (a) Manhattan-distance-based silhouette score, (b) distortio n277 3 n d ) s n All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 14 or sum of s quared erro rs, and (c ) even ness represented by the difference in f r action between the 278 largest and smallest clusters. Based on these tests, we chose spectral clustering with K=5. 279 280 Based on these tests, we first elimina t ed DBSCAN because the inferred num ber of clusters 281 was 131, a far too complex model to be useful or interpretable. Next, we e l iminated hierarchic al 282 clustering because the sizes of the resulting clusters were more uneven th an the other methods. 283 Spectral clustering and k-means clustering were ultimately more difficult to choo se between, but 284 when we focused on the sh apes of the distortion curves across the va lues of K, we observed that k-285 means lacked an “ elbow” to show a clear optimal K va lue wh ereas spec tral clus tering clearly 286 indicated 5 as an idea l K with a local minimum. Thus , we chose spectral cl u stering with K=5 as our 287 unsupervised subtyping model. The si zes of the final clust ers were: (1) 441 - 11%, (2) 686 - 17%, 288 (3) 1,151- 28%, (4) 796 - 20%, and (5) 1,004 - 25%. Figure 2 ill ustrates t he eigenvectors of the 289 affinity matrix which were used fo r clustering the data points. 290 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 1 5 291 Figure 2: pairwise scatter plots of the first five eigenvectors of the affinity matrix used for spectral 292 clustering, colored by cluster. This five-dimensional eigenvector space was used for clustering. The 293 diagonal shows kernel density estimator plots for each of the five eigenvectors. 294 295 296 5 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 1 6 Data-Driven Cluster Characterization 297 After clustering, we aimed to characterize these clusters by observing patterns in clinica l298 presentation (prevalence) amongst the input features. We performed two sets of z-scor e299 proportion tests comparing prevalence of each feature between each cluster and the other fou r300 clusters in our training set. The first set of tests was performed on the original cluster derivatio n301 cohort, and the features included were the 17 input features (symptoms and comorbidities wit h302 prevalence > 5%) as well as ICD-defined anatomical subtypes of endometriosis (Figure 3). 303 304 Figure 3: feature tests for the non-genotyped PMBB training set. Shown are (a) z-scores for the 305 difference in proportion tests, annotated with p-values that are significant and (b) feature 306 prevalence by cluster to provide context for the z-score tests. 307 308 Among the five clusters identified in the training set, there were many input features an d309 ICD-based anatomical subtypes with significantly different proportions. To identify distinguishin g310 6 l e r n h d g All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 17 features between the clust ers, we focu s on phenotypes which were signi fica ntly enriched and h ad 311 the highest prevalenc e in that cluster. Cluster one had the highest rates of (and was signi ficantly 312 enriched for) dysuria (Z=8.9), migraine (Z=10.6), IBS (Z=10.3), fibromyalgia (Z=15.3), asth ma 313 (10.3), abdominal pelvic pain (Z=13 . 6), and shortness o f b reath (Z=13.5). Cluster two had the 314 highest rates of the following signif icantly enriched traits: dysmenorrhea (Z=21.9), infertilit y 315 (Z=5.9), irregular menstruation (Z =31.75), leiomyoma of uterus (Z =21.9), and ut erine 316 endometriosis defined by IC D-9 617.0* or I CD-10 N80.0* (Z=13.4). Cluster thr ee’s defining features 317 were high risk pregnancy supervision ( Z=7.1), superficial lesions defined by ICD-9 617.3* or ICD -10 318 N80.3* (Z=7 .1), and lower abdomin al pain (Z=14. 6). Individ uals in cl uster four had highest 319 prevalence of abnormal cholesterol (Z=33.1) and hyp ertension (Z=33.9), whi le clust er five was only 320 enriched for unspecified endometrio sis defined as ICD -9 617.9* or I CD-10 N8 0.9* (Z=7.0). 321 The second set o f tests wa s perf ormed on a sub set of endometriosis cases (N=682) from the 322 genotyped PMBB for whom chart reviews were performed by OB-GYN clinical fel lows at th e 323 University of Pennsylvania Hospital System. The feat ures tested were gold standard confirme d 324 diagnoses (endometriosis, adenomyo s is, fibr oids, and any I CD false po sitives), surgical subtypes, 325 hormone use at the time of con firmation procedure, and symptoms identifie d from a combination 326 of structured data and notes (Figure 4 ) . 327 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 1 8 328 Figure 4: feature tests for the chart reviewed PMBB dataset. Shown are (a) z-scores for the 329 difference in proportion tests, annotated with p-values that are significant and (b) feature 330 prevalence by cluster to provide context for the z-score tests. 331 Because the size of our chart-reviewed dataset was limited, there were fewer significan t332 tests. For cluster one, the phenotypes which were most significantly prevalent were interstitia333 cystitis (Z=3.8) and fibromyalgia (Z=6.9). For cluster two, the defining features were confirme d334 adenomyosis status (Z=3.7), confirmed uterine fibroids (Z=7.1), and symptomatic bleeding (Z=5.3)335 Cluster three’s most highly enriched features were pelvic pain (Z=3.5) and hormone use at the tim e336 of surgery (Z=4.1). Considering the enriched features for each cluster among the two sets of tests337 we defined the following labels for 5 clusters: (1) pain comorbidities, (2) uterine disorders, (3 )338 pregnancy complications, (4) cardiometabolic comorbidities, and (5) EHR-asymptomatic. 339 8 t l d . e , ) All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 19 Candidate Gene Association Testing Stratified by Phenotypic Cluster 340 We applied the subtype classifications observed in our derivation set to our fo ur genetic 341 association datasets, PMBB, eMERGE, AOU, and UKBB. We used a K-nearest n eighbors model with 342 the same 17 EHR-derived features to a ssign endometrio sis cases to the five p henotypes (Table 2). 343 Table 2: c ounts and pr opo rtions of en dometriosis cases in each cluster by dataset. 344 Datase t P a i n Comorbidities Uterine Disorders Pregnancy Complications Cardiometabolic Comorbidities EH R - Asymptomatic Cluster Derivation Set: Training 441 (10.8%) 686 (16.8%) 1,151 (28.2%) 796 (19.5%) 1,004 (24.6%) Genetic Association Sets: AOU 713 (21.8%) 690 (21.1%) 723 (22.1%) 783 (23.9%) 362 (11.1%) eMERGE 495 (22.1%) 505 (22.5%) 382 (17.0%) 709 (31.6%) 152 (6.8%) PMBB 200 (16.7%) 222 (18.5%) 273 (22.8%) 366 (30.6%) 137 (11.4%) UKBB 231 (5.1%) 607 (13.4%) 842 (18.5%) 285 (6.3%) 2,576 (56.7%) Meta-Analysis Totals: META 1,639 (14.6%) 2,024 (18.0%) 2,220 (19.7%) 2,143 (19.0%) 3,227 (28.7%) 345 The smallest clust er was the pain com o rbidities cluster, with only 14.6% of total 346 endometriosis cases being assigned t o this cluster. The EHR-asymptomatic cl uster was the largest 347 cluster overall. The other three cluster s occurred i n relatively even proportio ns in the overall 348 meta-analysis group at 18.0% (uterine disorders), 19.7% (pregnancy complic ations), and 19.0% 349 (cardiometabolic como rbidities). 350 To establish a reference for the expected leve l of signal replication, we began with a positive 351 control test. We conducted associatio n tests on 39 established gene tic locat ions (autosomes only) 352 known t o be linked to endometrio sis. ( Figure 5). 353 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 2 0 354 Figure 5: results for our endometriosis case vs control positive control association tests at each of 355 the 39 known loci. Shown are the lead SNPs from the Rahmioglu et al 2023 GWAS as well their tag 356 SNPs in LD (kb distance 0.1). X-axis labels are from the known GWAS. 357 358 Our positive control test resulted in fourteen replicating loci. Only one was genome-wid e359 significant, RNLS/10q23 .31 (P = 1.91x10 -9 , rs792212:T). Thirteen were significant at a Bon Ferroni -360 corrected threshold of 0.05 / 39: WNT 4/1p3 6.1 2 (P = 9.12x10 -8 , rs2235529:T), DNM 3/1q 24.3 ( P =361 3.54x10 -5 , rs655853:C), GREB1/2p25 .1 (P = 6.55x10 -4 , rs34532804:A), PDLI M5/4q22. 3 (P = 8.58x10362 4 , rs1493112:T), EBF1/5q33. 3 (P = 7.64x10 -4 , rs1878936:C), SYNE1 /6q 25.1 (P = 3.20x10 -4363 rs13206045:C), GDAP1/8q21.11 (P = 4.19x10 -7 , rs10957712:T), CDKN2B-AS 1/ 9p21.3 (P = 1.75x10 -6364 rs10122243:T), ASTN2/9q3 3.1 (P = 9.01x10 -4 , rs62576127:A), ABO/9q34.2 (P = 5.87x10 -4365 rs495828:G), FSHB/1 1p14 .1 (P = 1.17x10 -3 , rs11031006:A), WT1/11p 14.1 (P = 2.85x10 -4366 rs72638188:T), DLEU 1/13q14 .2 (P = 2.24x10 -4 , rs9568417:G). 367 To test whether stratifying by clinical presentation allowed for greater resolution in geneti c368 associations, we performed case-control candidate gene association studies for the five phenotypi c369 clusters by meta-analyzing ancestry-stratified summary statistics from four EHR-linked geneti c370 datasets: PMBB, eMERGE, AOU, and UKBB. We observe 18 / 39 loci (46%) significantly associatin g371 with one or more clusters (Figure 6a, 6b). Also, for up to 15 loci, the cluster-stratified phenotype s372 0 e - = - , , , , c c c g s All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 2 1 yield stronger associations than the positive control despite having smaller sample sizes (Figur e373 6c). 374 375 Figure 6: Phenotype-specific association test results. The top panel (a) indicates which known loci 376 were significantly replicated by the positive control and five clusters. The bottom left panel (b) 377 shows the number and names of statistically significant associations for each phenotype. The 378 bottom right panel (c) shows the number of loci for which each phenotype had a more significant 379 association than the baseline. 380 The smallest clust er, cluster one, with high rates of pain comorbidities, was not significantl y381 associated with any known loci, but it was more significantly associated than the positive contro l382 for eight loci as shown in Figure 6c. The uterine disorders cluster (two) was significantly associate d383 with four loci, WNT4/1 p36. 12 , DNM 3/ 1q24.3 , GREB1 /2p2 5.1 , and GDAP1 /8q 21.11 . Out of the seve n384 1 e y l d n All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 22 loci significantly associated with the p regnancy complications clust er (three ), three of th em were 385 not signi ficantly associated with any o ther clusters or the positive contr ol: K DR/4q1 2 , 7p1 5.2 , and 386 KCTD9/8 p21. 2 . Cluster f our, enrich ed for cardiometabolic c omor bidities, was sig nificantly 387 associated with one locus, RNLS/10q2 3.31 , the strongest hit from the positi ve control. RNLS w a s 388 also significantly associated with clust ers three and five. El even loci were si gnificantly associated 389 with the EHR-asymptomatic clus ter, and six o f those ( BSN /3p2 1.31 , ID 4 /6p22 .3 , CD 109 /6q13 , 390 MLLT10/ 10p1 2.31 , IGF1/ 12q23 .2 , and SKAP1/17q 21.3 2 ) had no other associ ations, even with the 391 positive control. 392

393 Endometriosis presents with hetero geneous symptoms ranging fr om s evere pain to 394 infertility, contributing to varying pa tient experiences and treatment responses. Several large 395 genome-wide ass ociation studies and meta-analyses have been performed for endometriosis to-396 date. However, the genomic underpi nnings of endometriosis remain inco mplete ly understood, 397 largely due to the clinica l het erogeneity and the limita tions of traditional genome-wide association 398 studies (GWAS) that aggregate all cas es into a single analysis pool. This a pproach may obscure 399 genetic variations specific to different endometriosis phenotypes, thus necessitating more re fined 400 stratification techniques. There are various approaches to phenotyping participants for these 401 studies including surgical notes and electronic health records. While there h ave also been analyses 402 which account for disease pr ogres sion (17), there have not been any genom e-wide investigations 403 into the genetics underlying the heterogeneous presentation patterns of endo metriosis. 404 In this stu dy, we aimed to investiga t e the genetics of het erogeneity in en dometriosis by 405 defining data-driven subtypes in wom en from the non -genotyped PMBB end ometriosis population 406 (N=4,078) . We extrac ted c linical features known to be associated wit h endometriosis and 407 performed unsupervised spectral clust ering, identifying five clusters. Unsupe rvised clustering was 408 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 23 an ideal approach for this stud y becau se it a way to find pat terns in the d at a without introducing 409 prior kn owledge or bias. We cho se spe ctral clustering with five clusters based on empirical metrics 410 measured by comparing four diffe rent unsupervised clustering methods acro ss a range o f K values 411 (2-20 clusters). This method h ad th e best tradeoff between s quared error, silhouette score, and 412 cluster evenness. 413 To understand the c linical presentatio n patterns of each of the c lusters, we compared the 414 rates of the inpu t feat ures, diagnoses, and chart-reviewed p henotypes am ongst them. Based on 415 statistical enrichment testing across the features, the clust ers were labeled as (1) pain 416 comorbidities, (2) uterine disorders, (3) pregnancy complications, (4) cardiometabolic 417 comorbidities, and (5) EHR-asymptomatic. This nuance d phenotyping, which diverges from 418 traditional classifications, allows for a deeper understanding of the pathophysiological variations 419 within endometriosis and highlights the necessity of tailored therapeutic approaches. 420 After deriving and characterizing the clusters in the non-genotyped PMBB, we used a k-421 nearest neighbors’ model to trans fer the subtypes to the other four EHR-li nked genetic datasets, 422 PMBB, eMERGE, AOU, and UKBB. We performed ancestry-stratified candidate gene testing for each 423 of the clusters using SAIGE and ide ntified eight genome-wide significant signals. The gen etic 424 analysis of these clusters yiel ded intr iguing results. While 4 6% of previou sly known GWA S loci 425 were replicated in our study, signific ant differences in loci association s ac ross the clusters were 426 observed. For instance, genes like WN T4 and GREB1 showed specific associations with the uterine 427 disorders and EHR asymptomatic clus ters, suggesting that these genes migh t play distinc t roles in 428 the pathogenesis of these phenotypic presentations o f endometrio sis. C on versely, th e BSN gene, 429 although not statisti cal ly significant, d emonstrated greater significance in the pain and pregnancy 430 complications clusters, indicating a possible link to neurovascular or in flammatory mechanisms 431 that could exacerbate these conditions. 432 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 24 Renalase (RNLS) is the protein associated with our only genome-wi de significant 433 association fr om the positive control, RNLS/10q 23.3 1 . At the Bonferroni si gn ificance threshold, the 434 association with RNLS was si gnifi cant for three out o f five sub-ph enotypes: pregnancy 435 complications, cardiometabolic comorbidities, and EHR-asymp tomatic. It was the onl y significant 436 association with the cardiometabolic cluster. RNLS is highly expressed in the heart and contributes 437 to regulating blood pressure (43). I n genetic association studies, RNLS has been previously 438 associated with t ype 1 diabetes (44) a nd smoking initiation (45). Smoking is a known risk factor o f 439 endometriosis. 440 Cluster three, with high rates of pre gnancy-related complicat ions such as infertility and 441 high-risk pregnancy, was sig nificantly associated with seven loci including FSHB (P = 1.8 x 10 -4 ). 442 The FSHB gene codes fo r the beta-subunit of follicle-stimulating hormone (FSH). FSH is essential for 443 female fertilit y and has been shown to regulate m yometrial contracti le ac tivity (46). FSHB w a s 444 significantly associated with the positi ve control as well, but not with any of t he other clusters. 445 Cluster five, whic h was large ly asymp t omatic in th e EHR, was the largest c lu ster. Over half 446 (56%) of UK BB endometriosi s patients were assigned to this cluster . It is possible that those 447 assigned to this cluster fr om any dataset have symptoms that were not reco rded in the structured 448 data which we had access to. Two well-known endometriosi s loci from both o f the last major 449 GWASs are SY N E 1 and CDKN2B -AS1 ( 16,17), both of which were significant ly associated with the 450 positive control and the EHR-asymptomatic c luster. Six loci were associated with this clust er and no 451 other phenotypes: BSN , ID4 , CD109, MLLT10 , IGF1, and SKAP1. MLLT10 and BSN hav e been 452 previously associated with pain perc eption and maintenance (17) . Serum levels of IGF-1 are 453 significantly elevated in women with e ndometriosis (47). Gene e xpressi on of ID4 is down -regulated 454 in eutopic and ectopic endometria l tis sue of women with endometriosi s (4 8). CD109I and SKAP1 455 have been previously associated with e ndometrial cancers (49,50). 456 457 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 25 Subtyping complex diseases, like endo metriosis, is crucial fo r advancing prec ision medicine. 458 The findings f rom our study underscore the uti lity of EHR as a rich resource for disease subtyping 459 and genetic rese arch. T he linkage of detaile d clinic al data with gene tic inf ormation enables the 460 identification of phenotype-genotype c orrelations that are often dilu ted in broader GWAS analyses. 461 Furthermore, the use of spectral cluste ring helps elucidate t he het erogeneity within endometriosis, 462 providing a framework f or understanding the m ultiface ted na ture of the dise ase and facilit ating t he 463 development of personalized medicine. 464 However, it is essential to ackno wledge the limita tions of our study. One significant 465 constraint was the sample size, whic h was particularly li mited for some of th e smaller cl usters and 466 for individuals of non- European ancest ry. This limitation could potentially int roduce bias and af fect 467 the generali zability of our findings. Additionally , our study re lies on struct ured elec tronic healt h 468 data only, which m ay not capt ure th e full clinic al pic ture and cou ld be subject to inaccuracies or 469 incomplete records. Lastl y, this gene tic association analyses in this stud y only focused on the 470 candidate genes that are previously known to be as sociated with endometriosis. This appr oach 471 might hav e restricted our ability to discover novel genetic loci poten tial ly r elevant to th e specific 472 clusters identified . Despite these li mitations, our stud y marks a meani ngful advancemen t in 473 understanding the genetic factors that may contribute to t he hetero geneity observed in 474 endometriosis. By focusin g on genetic associations gleaned fr om electronic health records, we offer 475 a novel perspective tha t could be instrumental in future research and treat ment approaches. To 476 expand upon the current findings, future research should aim to perform comprehensive GWAS 477 across the identified endometriosis subtypes. This will enable the de tection of novel loci that could 478 be crucial fo r understanding the di stinct mechanisms underlying each subtype. Additionall y, 479 integrating multi-omics data (such a s transcriptomic, proteomic, and metabolomic data) could 480 further refine the molecu lar signatu res associated with each c luster, en hancing the biological 481 interpretability of the genetic associat ions. Another p romising avenue is the longitudinal study of 482 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 26 these cl usters to assess disease progr ession and treatment outcomes, whic h could inform more 483 effective, personalized therapeu tic strategies. 484 In conclusion, our research highlig hts th e importance of subtype-specific studies in 485 elucida ting the genetic basis of endometriosis. By leveraging t he c apabilities of EHR-linked 486 biobank s and employing advanced clu stering techniques, we pav e t he wa y for more targeted an d 487 effective approaches to understanding and managing this complex di sease. 488

489 Research reported in this publicatio n was supported by the Eunice Kennedy Shriver National 490 Institute of Child Health and Hu man Development of the National Instit utes of Health under award 491 number R01HD110567. 492 493 We acknowledge t he Penn Me dicine BioBank (PMBB) f or providing data a nd thank the pati ent-494 participants of Penn Medicine who c o nsented to participate in this research program. We w ould 495 also like to thank the Penn Medicine BioBank team and Regeneron Genetics Center fo r providing 496 genetic variant data for analysis. T he PMBB is appr oved under IRB p rotocol# 813913 and 497 supported by Perelman School o f Me dicine at University of Pennsylvania, a gift fr om the Smilow 498 family, and t he National Center for Advancing Translational Sciences of th e Na tional Insti tutes of 499 Health under CTSA award numbe r UL1TR001878. 500 501 This phase of the eMERGE Network w as initiated and funded by the NHGRI through the following 502 grants: U01HG008657 (Group Health Cooperative/University of Washington); U01HG008685 503 (Brigham and Women’s Hospital); U01HG0086 72 (Vanderbilt Univer sity Medica l Center); 504 U01HG008666 (Cincinnati Chil dren’s Hospital Medical Cent er); U01HG0 06379 (Mayo Clinic); 505 U01HG008679 (Geisinger Clinic); U01HG008680 (Columbia Universit y Heal th Scienc es); 506 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 27 U01HG008684 (Chil dren’s Hospital of Philadelp hia); U01HG008 673 (Nort hwestern University); 507 U01HG008701 (Vanderbilt Uni versit y Medical Center serving as the Coo rdinating Center) ; 508 U01HG008676 (Partners Hea lthcare /Broad Institute); and U 01HG00866 4 (Baylor College of 509 Medicine). 510 511 We gratefully acknowledge A ll of U s participants for their contributions, without whom this 512 research would not have been p ossi ble. We also thank the National Institu te s o f Health’s All of Us 513 Research Program for making available the participant data examined in this study. 514 515 This research has been conducted u sing the UK Biobank Resource unde r Application Number 516 32133. 517

518 1. Gao X, Outley J, Botteman M, Spalding J, Simon JA, Pashos CL. Ec onomic burde n of 519 endometriosis. Fertil Steril. 2006 Dec 1;86(6):1561–72 . 520 2. Burney RO, Giudice LC. Pathogenesis a nd pathophysiology o f endometrio sis. Fertil Steril. 2012 521 Sep 1;98(3):511–9. 522 3. Greene R, Stratton P, Cleary SD, Ballweg ML, Sinaii N. Diagno stic experience a mong 4,334 523 women repo rting sur gically diagnosed endometriosis. Fertil Steril. 2009 Jan;9 1(1):32–9. 524 4. As-Sanie S, Soliman AM, Evans K, Erpel ding N, Lanier RK, Katz NP. Short-actin g and Long -acting 525 Opioids Utilization among W omen Dia gnosed with E ndometriosis in the United States: A 526 Population-based Claims Study. J Minim Invasive Gynecol. 2021 Feb 1;28(2): 297-306.e 2. 527 5. Soliman AM, Surrey E, Bonafede M, Nelson JK, Castelli-Haley J . Real-World Evaluation of Direct 528 and Indirect Economic Burden Amo ng Endometriosis Patients in the United States. Adv Ther. 529 2018 Mar 1;35(3):408–23. 530 6. Shakiba K, Bena JF, McGill KM, Minger J, Falcone T. Surgical Treatment of Endo metriosis: A 7-531 Year Follow-up on the Requirement for Further Surgery. Obstet Gynecol. 2008 532 Jun;111(6):1285. 533 7. Singh SS, Suen MWH. Surgery f or end o metriosis: bey ond medical therapies. F ertil Steril. 2017 534 Mar 1;107(3):549–54. 535 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 28 8. Ellis K, Munro D, Clarke J. Endometrio s is Is Undervalued: A Call to Action. Fr o nt Glob Womens 536 Health [Internet]. 202 2 [cited 2022 Dec 1];3. Available from: 537 https://www.frontiersin.or g/articles/10.3389 /fgwh.2022. 902371 538 9. Simoens S, Dunselman G, Di rksen C, Hummelshoj L, Bokor A, B randes I, et al. The burden o f 539 endometriosis: c osts and quality of life of women with endometriosis and trea ted in referral 540 centres. Hum Reprod Ox f Engl. 2012 M ay;27(5):1292–9 . 541 10. Penrod N, Okeh C, Velez Edwards DR, B arnhart K, Senapati S, Verma S S. Leveraging electronic 542 healt h record data fo r endometrio sis r esearch. Front Digit Health [Internet]. 2023 Jun 5 [cited 543 2024 Apr 12];5. Available from: https://www.fr ontiersin.org/j ournals/digital-544 healt h/articl es/10.3 389/fdgth. 2023 .1 150687/ful l 545 11. Soliman AM, Fuldeore M, Snabes MC. F actors Ass ociated with Time to Endom etriosis Diagnosi s 546 in the United States. J Womens Health. 2017 Jul;26(7):788 –97 . 547 12. Becker CM, B okor A, Heikinheimo O, H orne A, Jansen F, Kiesel L, et al. ESHRE guideline: 548 endometriosis†. Hum Reprod Open. 2022 Jan 1;2022(2):hoac009. 549 13. Wykes CB, Clark TJ, Khan KS. Accuracy of laparo scopy in the diagnosis o f endo metriosis: a 550 systematic quantita tive review. BJOG Int J Obstet Gynaecol. 2004 Nov;111(11):1204–12. 551 14. Nisolle M, Paindav eine B, Bourdon A, B erlière M, Casanas-Roux F, Don nez J. Histologic study o f 552 peritoneal endometriosis in infertile w omen. Fertil Steril. 1990 Jun;5 3(6):984–8. 553 15. Fauconnier A, Fritel X, Chapron C. [End ometriosis and pelvic pain: epidemiological evidence of 554 the relationship and implications]. Gynecol Obstet Fertil. 2009 Jan;37(1):57– 6 9. 555 16. Sapkota Y, Steinthorsdottir V, Mor ris A P, Fassbender A, Rahmioglu N, De Vivo I, et al. Me ta-556 analysis identifies five n ovel loci associ ated with endometriosis highlighting key genes involved 557 in hormone metabolism. Nat Commun. 2017 May 24;8(1):15539. 558 17. Rahmioglu N, Mortlock S, Ghiasi M, Møller PL, Stefansdottir L, Galarneau G, et al. The genetic 559 basis of endometriosis and com orbidit y with other pain and in flammatory co nditions. Nat 560 Genet. 2023 Mar;55(3):423–36 . 561 18. Saha R, Pettersson HJ, Svedberg P, Olov sson M, Bergqvist A, Marions L, et al. Heritability of 562 endometriosis. Fertil Steril. 2015 Oct 1;104(4):947–52 . 563 19. Lee SH, Harold D, Nyholt DR, Goddard ME, Zondervan KT, Williams J, et al. Estimation and 564 partitioning o f polygenic variation captured by common SNPs for Alzheimer’s disease, multip le 565 sclerosis and endometriosis. Hum Mol Genet. 2013 Feb 15;22(4):832–41 . 566 20. Brawn J, Mo rotti M, Zondervan KT, Bec ker CM, Vincent K. Central changes ass ociated with 567 chronic pelvic pain and endometriosi s. Hum Reprod Update. 2014;20(5):737 – 47. 568 21. Nezha t C, Li A, Abed S, Balassiano E, Sol iemannjad R, Nezha t A, et al. Strong As sociation 569 Between Endometriosis and Symptom atic Leiomyomas. JSLS. 2016;20(3):e20 16.0005 3. 570 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 29 22. Verma A, Damrauer SM, Naseer N, Weaver J, Kripke CM, Guare L, et al. The Penn Medicine 571 BioBank: To wards a Genomics -Enabled Learning Healthcare System to Accelerate Precision 572 Medicine in a Diverse P opulation. J Pers Med. 2022 Dec;12(12):1974. 573 23. Gottesman O, Kuivaniemi H, Tromp G, Faucet t WA, Li R, Manolio TA, et al. The Electronic 574 Medical Records and Genomics (eMER GE) Network: past, present, and future. Genet Med Of f J 575 Am Coll Med Genet. 2013 Oct;15(10):761–71 . 576 24. GWAS and enrichment analyses of non -alcoholic fatty liver disease identify new trait-associated 577 genes and pathways acros s eMERGE Network | Sprin gerLink [Internet]. [cited 2023 Aug 14]. 578 Available from: https://link.springer.com/article/10 .1186/s12 916-019-136 4-z 579 25. Data Bro wser | All o f Us Public Data Br owser [Internet]. [cited 2024 Mar 30]. Available from: 580 https://databrowser.researchallofus.org/ 581 26. Genomic data in the All of Us Research Program. Nature. 2024;62 7(8003):340 –6. 582 27. The “All of Us” Research Pr ogram. N En gl J Med. 2019 Aug 15;381(7):668–76 . 583 28. Sync For Science [Internet]. [cited 202 4 Mar 30]. Available from: https://syncfor.science/ 584 29. Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp K, et al. The UK Bi ob ank resource 585 with deep phenotyping and genomic d ata. Na ture. 2018 Oct; 562(7726):203 – 9. 586 30. Hallinan CM, Ward R, Hart GK, Sullivan C, Pratt N, Ng AP, et al . Seamless EMR data access: 587 Integrated governance, digital hea lth a nd the OMOP-CDM. BMJ Health Care Inform. 2024 Feb 588 21;31(1):e100953 . 589 31. An integrated map of genetic variation from 1,092 human genomes. Nature. 2012 Nov 590 1;491(7422):56–6 5. 591 32. Jia H, Ding S, Xu X, Nie R. The latest res earch progress on spectral clustering. Neural Comput 592 Appl. 2014 Jun 1;24(7):1477– 86. 593 33. Damle A, Minden V, Ying L. Simple, direct and efficient multi-way spectral clustering. Inf 594 Inference J IMA. 2019 Mar 15;8(1):181–203. 595 34. Khan K, Rehman SU, Aziz K, Fon g S, Sar asvady S. DBS CAN: Past, present and fu ture. In: The Fifth 596 International Conference on the Applications o f Digital Information and Web Technologies 597 (ICADIWT 2014) [Internet]. 2014 [cited 2024 Mar 30]. p. 232–8. Avail able fro m: 598 https://ieeexplore.ie ee.org/abstract/d ocument/68146 87 599 35. Algorithms fo r hierarchical clustering: an overview - Murtagh - 2012 - WIREs Data Mining and 600 Knowledge Di scovery - Wiley Online Li brary [Internet]. [cited 2024 Mar 30]. Available from: 601 https://wires.onlinelibrary.wiley .com/ doi/full/10.1 002/widm. 53 602 36. Kanungo T, Mount DM, Netanyahu NS, Piatko CD, Silverman R, Wu AY. An ef fic ient k-means 603 clustering algorithm: analysis and implementation. IEEE Trans Pattern Anal Mach Intell . 2002 604 Jul;24(7):881 –92. 605 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 30 37. Watts V. 9.5 Statistica l Inference fo r T wo Population Pr oportion s. 2022 Sep 1 [cited 2024 Apr 606 16]; Available from: https://ecampusontario.pressbo oks.pub/introstats/chapter/9-5-607 statistical-inference-for-two -population-propo rtions/ 608 38. Chang CC, Chow CC, Tellier LC, Vattikut i S, Purcell SM, Lee JJ. Second-generation PLINK: rising to 609 the chal lenge of larger and richer datas ets. GigaScience. 2015 Dec 1;4(1):s13742-015-0047– 8. 610 39. Islam MJ, Wu QMJ, Ahmadi M , Sid-Ahmed MA. Investigating the Performance of Naive- Bayes 611 Classifiers and K - Nearest Neighbor Classifiers. In: 2007 International Con ference on 612 Convergence Inf ormation Technology ( ICCIT 2007) [Internet]. 2007 [cited 2024 Apr 16]. p. 613 1541–6 . Available from: https://ieeexplore.ieee.org/abstract/document/ 442 0473 614 40. Zhou W, Nielsen JB, Fritsche LG, Dey R, Gabrielsen ME, W olford BN, et al. Effici ently controlling 615 for case-contr ol imbalance and sample relatedness in large-scale genetic assoc iation studies. 616 Nat Genet. 2018 Sep;50(9):1335 –41. 617 41. Purcell S, Nea le B, Todd-Br own K, Tho mas L, Ferreira MAR, Bender D, et al. PLINK: A To ol Set 618 for Whole-Genome Ass ociation and Po pulation-Based Linkage Analyses. Am J Hum Genet. 2007 619 Sep;81(3):559–75. 620 42. Borenstein M, Hedges LV, Higgin s JPT, Rothstein HR. A basic intr oduction to fi xed-effect and 621 random-effects models f or meta-analys is. Res Synth Methods. 2010;1(2):97–1 11. 622 43. Heydarpour M, Parksook WW, H opkins PN, Pojoga LH, Williams GH, Williams JS. A candidate 623 locus in the renalase gene and suscepti bility to blood pressure respon ses to the dietary salt. J 624 Hypertens. 2023 May;41(5):723 . 625 44. Onengut-Gumuscu S, Chen WM, Burren O, Co oper NJ, Quinlan AR, Mychaleck yj JC, et al. Fine 626 mapping of type 1 diabetes susceptibility loci and evidence for colocalization of causal variants 627 with lymphoid gene enhancers. Nat Genet. 2015 Apr;47(4):381–6 . 628 45. Saunders GRB, Wang X, Chen F, Jang SK , Liu M, Wang C, et al. Genetic diversity fuels gene 629 discovery f or to bacco and alcohol use. Nature . 2022 Dec;612(7941):720– 4. 630 46. Stilley JAW, Segaloff DL. FSH Actions a nd Pregnancy: Looki ng Beyond Ovaria n FSH Receptors. 631 Endocrinology. 2018 Dec 1;159(12):4033–42 . 632 47. Heidari S, Kolahdouz-Mohammadi R, K hodaverdi S, Tajik N, Delbandi AA. Expr ession levels of 633 MCP-1, HGF, and IGF-1 in end ometriotic patients compared with non -endome triotic controls. 634 BMC W omens Health. 2021 Dec 20;21(1):422. 635 48. Amirteimouri S, Ashini M, Ramazanali F, Aflatoonian R, Af sharian P, Shahhoseini M. Epigenetic 636 role of the nuclear factor NF-Y on I D ge ne family in endometrial tissues o f wo men with 637 endometriosis: a case contr ol study. Reprod Bi ol Endoc rinol. 2019 Mar 15;17( 1):32. 638 49. Al-kuraishy HM, Al-Maiahy TJ, Al-Garee b AI, Alexiou A, Papadakis M, Saad HM, et al. The 639 possible r ole furin and furin inhibito rs in endometrial adenocarcinoma: A nar rative review. 640 Cancer Rep. 2024;7(1):e1920. 641 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint 31 50. Kho PF, Wang X, Cuéllar-Partida G, Dör k T, Go ode EL, Lambrechts D, et al. Multi-tissue 642 transcriptome-wide ass ociation study i dentifies eight candidate genes and tiss ue-specific gene 643 expression underlying endometrial cancer susceptibility. Commun Biol. 2021 Oct 21;4(1):1–8. 644 645 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.24306092doi: medRxiv preprint