Genome-wide meta-analyses of cross substance use disorders in diverse populations
Department of Medical and Molecular Genetics, Indiana University School of Medicine, Indianapolis, IN 46202 USA
Department of Psychology and Program in Neuroscience, Middlebury College, Middlebury, VT 05753 USA
Department of Genetics and Genome Sciences, Case Western Reserve University, Cleveland, OH 44106 USA
Computer Science and Artificial Intelligence Lab, Massachusetts Institute of Technology, Cambridge, MA 02139 USA
Department of Biostatistics and Heath Data Science, Indiana University School of Medicine, Indianapolis, IN 46202 USA
Department of Biochemistry and Molecular Biology, Indiana University School of Medicine, Indianapolis, IN 46202 USA
Abstract
Substance use disorders (SUDs, including alcohol, cannabis, opioids, and tobacco) represent significant public health challenges. The estimated heritability of SUDs is ~50% and many individuals experience multiple SUDs concurrently. Studies have demonstrated the existence of genes shared across multiple SUDs, and identifying these SUD-shared genes is critical to developing novel prevention and treatment strategies. Here, we conducted the largest cross SUD meta-analysis to date to identify SUD-shared genes using samples genetically similar to 1000 Genomes Project European (1kg-EUR-like), African (1kg-AFR-like), and American mixed (1kg-AMR-like) populations. We defined variants that had the same direction of effects across different SUDs (i.e., concordant variants) as SUD-shared. In total, we identified 220 loci, including 40 novel loci that were not reported as SUD-associated in previous genome-wide association studies. Through gene-based analyses, gene mapping, and gene prioritization, we identified 785 SUD-shared genes. These genes are highly expressed in the amygdala, cortex, hippocampus, hypothalamus, and thalamus; and are primarily highly expressed in neuronal cells, suggesting that more brain regions may be involved in SUDs than previously reported. Concordant variants explained 56–96% of the SNP-heritability of each SUD in the 1kg-EUR-like sample. Furthermore, the top 10% of individuals in the 1kg-EUR-like and 1kg-AMR-like samples with the highest polygenic scores had odds ratios ranging from 1.95–2.87 to develop SUDs, and these polygenic scores could potentially be used to identify high-risk individuals. Lastly, using a real-world dataset, we identified seven SUD-shared genes targeting drugs that may be repurposed for treating SUDs, particularly in those suffering from comorbid SUDs.
Untitled section
Subject terms: Addiction, Genetics
Article notes
Untitled section
Received 2024 Dec 9; Revised 2025 Aug 28; Accepted 2025 Sep 23; Issue date 2026.
Introduction
Substance use disorders (SUDs) have devasting consequences on affected individuals, their families, and society. Globally, approximately 5.5% of disability-adjusted life-years are attributable to SUDs [1]. The estimated heritability of each SUD is ~50% [2] and identifying SUD-associated genes will not only help us understand the genetic etiologies of SUDs, but also facilitate the development of novel prevention and treatment strategies.
SUDs share common features such as uncontrolled use of substances and withdrawal/negative affect [3]. Many people suffer from multiple SUDs simultaneously with about one in four individuals diagnosed with SUD having two or more SUDs [4]. Furthermore, in individuals with an SUD diagnosis other than alcohol use disorder, 56.8 to 97.5% had at least one other comorbid SUD diagnosis as well [5]. Twin studies [6–8] and genetic correlation studies [9, 10] have confirmed the existence of genes shared across multiple SUDs. Identification of these SUD-shared genes can elucidate the common genetic pathways to SUDs and importantly, facilitate the development of novel and more effective treatment methods. This is particularly important for individuals with comorbid SUDs as they typically experience more severe mental and physical health challenges and require complex treatment approaches. One way to identify genes shared across multiple SUDs is to perform a univariate analysis by defining SUD cases as having any SUD and defining controls as not having any SUD. However, individual level genotype data is typically required, therefore studies using this method are often limited by small to moderate sample sizes [11–15], making replication difficult. Multivariate methods such as genomic structural equation modeling (genomic SEM) [16] and meta-analysis [17] can utilize summary statistics from large-scale genome-wide association studies (GWAS) and thus are more powerful. Genomic SEM derives a common risk factor by using LD (linkage disequilibrium) score regression [18] and has successfully identified multiple SUD-shared genes [9, 19]. However, model misspecifications and LD pattern mismatches among different studies can lead to results that cannot be replicated by other studies. Additionally, while findings are supposed to reflect genes shared across multiple SUDs, some studies report genes with opposite directions of effects across different SUDs [9]. For example, two variants identified using genomic SEM, rs2424952 and rs28567725, had discordant effects for problematic alcohol use and cannabis use disorder [9]. Although these discordant findings may still reflect shared genetic factors across SUDs, the underlying mechanisms are difficult to interpret. Furthermore, drugs developed targeting these discordant genes may be less effective in individuals with comorbid SUDs.
In this study, we performed the largest cross SUD meta-analysis to date to search for SUD-shared genes in diverse populations. We chose meta-analysis because it does not estimate a common risk factor and thus avoids model misspecifications and LD pattern mismatches. This in turn provides a more powerful approach [20–22]. Since many individuals suffer from comorbid SUDs simultaneously, it is less likely that SUD-shared variants increase the risk for one SUD but decrease the risk for another SUD. Therefore, we considered variants with the same direction of effects in different SUDs (i.e., concordant variants) as SUD-shared for easier interpretation and drug development purposes. Multiple approaches were used to map identified concordant variants to genes, then a series of steps were used to prioritize identified genes. We also performed gene enrichment analyses, estimated the SNP-heritability explained by concordant variants, and calculated genetic correlations. Polygenic scores (PGS) were calculated by using concordant variants and their predictive power for SUD risk was tested using independent datasets. Lastly, we identified drugs targeting prioritized SUD-shared genes and tested whether they could be repurposed to treat SUDs by using a large-scale real-world dataset.
Materials and methods
Ethics approval and consent to participate
All methods were performed in accordance with the relevant guidelines and regulations. This study was approved by the Indiana University Institutional Review Borad (protocol number: 1610935245). Informed consent was obtained from all participants at each study.
Gene-based analysis
MAGMA [33] was used to perform gene-based analyses. Since SUDs are psychiatric disorders, only genes (N = 33,761) having median transcripts per kilobase million (TPM) > 0 in at least one of 13 brain tissues from GTEx [34] were included in gene-based analyses. Within each population, the top χ2 statistic option in MAGMA [33] was used to perform the gene-based analysis. Cross-population gene-based analyses were performed by using the meta-analysis method in MAGMA [33] with gene-based analysis results from each population as the input files. Genes presented in the analysis results of only one population were excluded.
Mapping significant variants to genes
Three strategies were used to map significant variants to genes: 1) Positional mapping using FUMA [30], i.e., whether a significant variant was located within a gene ( ± 10 kb from transcription start and end sites). Gene annotations were obtained from ENSEMBL gene v102 [35] by using ANNOVAR [32]. 2) eQTL mapping using FUMA [30], i.e., whether a significant variant was an eQTL of a gene (cis-eQTL only). eQTL datasets from PsychENCODE [36] were used. 3): Chromatin interaction mapping using FUMA, i.e., whether a variant contacted a gene by chromosomal looping (within 2 Mb). Similar to our gene-based analysis, we limited our analysis to genes that are known to be expressed in at least one brain tissue from GTEx [34]. Genes mapped by any of these strategies were included in gene prioritization.
Gene prioritization
Some mapped genes may be physically close to SUD-shared genes but may not be biologically relevant to SUDs. Therefore, we prioritized mapped genes using a series of steps. First, we checked whether mapped genes were in SUD-related pathways (alcoholism, amphetamine addiction, cocaine addiction, morphine addiction, nicotine addiction, dopaminergic synapse, GABAergic synapse, glutamatergic synapse, and MAPK signaling) defined by the Kyoto Encyclopedia of Genes and Genomics (KEGG: https://www.genome.jp/kegg/). Given the fact that these pathways were derived from animal studies and different genes with similar functions may cause SUDs in humans, we also checked whether any of the prioritized genes belonged to the same gene families as those in KEGG SUD pathways, or if they directly interacted with genes in those pathways. Gene-gene interactions were obtained from the STRING database (https://stringdb.org/) [37]. Genes with high confidence interactions (i.e., a confidence score ≥0.7 or obtained from high-throughput laboratory experiments and previous knowledge in databases) were retained. Finally, we queried the GWAS catalog [38] and retained genes that were associated with psychiatric traits, brain measurements, or brain functions. Genes identified by any of these steps were prioritized.
Brain dissection and cell type enrichment analyses
Four enrichment analyses were performed to identify: 1) in which brain dissections prioritized genes were highly expressed, 2) in which brain dissections prioritized genes were expressed at low levels, 3) in which brain cell types prioritized genes were highly expressed, and 4) in which brain cell types prioritized genes were expressed at low levels. We used single-cell RNA sequencing data from 105 dissections and 461 cell types generated by the BRAIN Initiative cell census network [39]; and to be consistent with the original publications, we used the same naming system. For the first enrichment analysis, the mean and standard error (SE) of RNA expression across all dissections were calculated for each gene; then we tested whether the expression of that gene in a particular dissection was significantly higher than the mean expression (> mean + 1.645*SE; α = 0.05). Each gene in each dissection was labeled as either highly or not highly expressed. Then, for each dissection, we tested whether the frequency of highly expressed prioritized genes was significantly greater than the frequency of all other genes using the Fisher’s exact test. Benjamini-Hochberg false discovery rate [40] was used to correct for multiple testing. The same procedure was performed in the second enrichment analysis except we tested whether the gene expression was significantly lower than the mean expression (<mean - 1.645*SE; α = 0.05). The other two enrichment analyses used the same procedure but tested cell types.
SNP-heritability estimation and calculating SNP-heritability explained by concordant variants for each SUD
Complex Traits Genetics Virtual Lab (CTG-VL https://vl.genoma.io/data) was used to estimate SNP-heritability using LD score regression [18]. SNP-heritability explained by concordant variants was estimated by using the LD score regression model [18, 41] implemented in LDAK (v5.2) [42, 43] on the observed scale. In order to most accurately estimate LD patterns, we limited this analysis to the 1kg-EUR-like sample due to lack of access to individual level data for the 1kg-AFR-like and 1kg-AMR-like samples. To estimate SNP-heritability in 1kg-EUR-like sample, we made a European reference panel by including Finnish samples from 1 kg [29] following the instruction of LDAK [42, 43] as data from the FinnGen consortium were included in our analyses. SNP-heritability was estimated by 1) using all available variants in each SUD GWAS, and 2) limited to concordant variants only. The estimate in 2) divided by the estimate in 1) was the SNP-heritability explained by concordant variants for each SUD.
Calculating genetic correlations
CTG-VL was also used to perform genetic correlation analyses using LD score regression [18]. To accomplish this, every individual SUD GWAS was uploaded to CTG-VL to calculate genetic correlations between the different SUDs. All variants from each SUD GWAS were included. We also calculated genetic correlations between our cross SUD meta-analysis (concordant variants only) and other publicly available GWAS, which are available at CGT-VL. Significant correlations were determined by Bonferroni correction. Genetic correlation analyses were limited to the 1kg-EUR-like sample for the same reason as the SNP-heritability estimations.
Polygenic score analyses
Samples from the All of Us research program (AOU) [36] and the Indiana Biobank (IB) [44] were used as the target datasets to perform PGS analyses. Genetically determined 1kg-EUR-like, 1kg-AFR-like, and 1kg-AMR-like samples from AOU version 8 dataset were used; variants with a population-specific allele frequency >1% or a population-specific allele count >100 in any population were included. In the AOU dataset, SUD status was obtained from participants’ electronic health records (EHR), then refined and validated by the AOU Data and Research center. IB data processing and QC were performed as described previously [44, 45]. The IB dataset did not have a sufficient sample size for the 1kg-AMR-like sample, therefore PGS analyses were performed only in 1kg-AFR-like and 1kg-EUR-like samples. SUD status was also obtained from EHRs and determined based on ICD9/10 codes. In both AOU and IB datasets, SUD cases were defined as those individuals having any SUD (i.e., not limited to AUD, CUD, OUD, and TUD) and controls were defined as everyone else. For any related samples (as determined by identity by descent), we kept only one individual from each family by first selecting those having SUDs and/or higher call rates. In cases where the related individuals were all controls or had similar call rates, then we randomly selected one individual from each family.
For the 1kg-EUR-like sample, we used the 1kg-EUR-like meta-analysis results as the discovery dataset and European samples from 1 kg [29] as the reference panel to estimate posterior effect sizes of concordant variants using PRS-CS [46]. For the 1kg-AFR-like and 1kg-AMR-like samples, in addition to the meta-analyses results from 1kg-AFR-like and 1kg-AMR-like, respectively, 1kg-EUR-like results were also used as the discovery dataset to increase the statistical power. However, since variants with discordant effects across different populations decrease the statistical power, we only kept 1kg-EUR-like variants with concordant effects in the1kg-AFR-like or 1kg-AMR-like samples. Combined posterior effect sizes across different populations were estimated by using PRS-CSx (i.e., meta-analysis results from PRS-CSx) [47]. African (for 1kg-AFR-like) and American Mixed (for 1kg-AMR-like) samples from 1 kg [29] were used as the reference panels. In the AOU dataset, PGS were calculated as the sum of the numbers of alleles multiplied by the posterior effect sizes. If a genotype was not successfully sequenced, then it was imputed as mean allele counts based on the allele frequencies via PLINK [48, 49]. In the IB dataset, PGS were calculated as the sum of imputation dosages multiplied by the posterior effect sizes.
To identify high-risk individuals based on their PGS, we compared the top 10% of individuals with the highest PGS to the remaining 90%. We used the top 10% as this approximately represents the prevalence of SUDs in the AOU 1kg-EUR-like and 1kg-AMR-like populations. Additionally, since the prevalence of SUDs differs between males and females, sex-stratified analyses were performed. Logistic regression was used with sex (not used in sex-stratified analyses), age, and the first 10 principal components of genetic ancestries as covariates. For SUD cases, age was defined as the age at the first diagnosis; for controls, age was defined as the age at the last interview. An array indicator was also included as a covariate in the analyses of IB as IB data were genotyped on two arrays. We also meta-analyzed the AOU and IB results in 1kg-EUR-like and 1kg-AFR-like samples weighted by the inverse of standard error using Metal [17].
Drug repurposing
The Drug Gene Interaction Database (DGIdb, v4.2.0, https://www.dgidb.org/) [50] was queried to identify drugs targeting our list of prioritized genes. Only drugs approved by the Food and Drug Administration (FDA) were included. The Anatomical Therapeutic Chemical (ATC) codes of drugs were obtained from the KEGG: https://www.genome.jp/kegg/drug/). Repurposable drugs were selected based on four criteria: 1) have an ATC code N, 2) are not currently approved to treat any SUD, 3) are shown to be related to SUD treatment through a literature search [51–58], and 4) have comparators (i.e., drugs with the same ATC level 4 codes as drugs we identified but do not target any gene we identified). Since the identified drugs belong to four ATC level 4 classes (N03AX, N05AX, N06AA, and N06BA), four cohorts corresponding to the four ATC classes using the active comparator design were derived. Data were obtained from Optum’s de-identified Clinformatics® Data Mart Database, which is a longitudinal database derived from medical claims and pharmacy claims data of a large national insurer cohort (https://www.optum.com/business/life-sciences/commercial-analytics/managed-markets.html). In each cohort, individuals with pharmacy claims belonging to the corresponding ATC class were identified with the year of drug initiation between 2008 to 2021. All individuals were 21 years or older. The clock-0 was defined as the first exposure date of repurposable drug or comparator and individuals with >365 days of enrollment prior to the clock-0 were included. Individuals concurrently initiated on both repurposable drugs and comparators on clock-0 were excluded.
In the active comparator design, the “repurposable drugs” were those drugs identified and the comparators were the remaining drugs in the corresponding repurposable drugs’ ATC level 4 classes. The outcome was time from clock-0 to diagnosis of SUDs. Individuals without an outcome were censored on the earliest date of drug discontinuation (i.e., without exposure for >60 days), enrollment discontinuation, or concurrent of initial of repurposable drug and comparator. The R package comorbidity was used to define the outcome and comorbidities. Cox proportional hazard models were used to investigate the association between repurposable drugs and the outcome. Cox proportional models were adjusted for age at clock-0, sex, race, and comorbidities, and stratified by the year of drug initiation.
Results
Identification of SUD-shared loci
We identified 428 independent lead variants (184 loci) in the 1kg-EUR-like sample, two independent lead variants (2 loci) in the 1kg-AFR-like sample, and three independent lead variants (1 locus) in the 1kg-AMR-like sample (Supplemental Table 2–4). In the 1kg-EUR-like sample, there were 8,220 GWS variants, and among 184 loci identified, 155 loci had ≥2 GWS variants. Locus 1 in the 1kg-AFR-like sample was an intergenic variant without any LD support; therefore, it was likely a false positive and was excluded from further analysis. The locus identified in the 1kg-AMR-like sample included alcohol metabolism genes, which are associated with AUD [59]. The 1kg-AMR-like sample only had AUD and OUD GWAS, and it is likely that this finding was driven by AUD. However, we included this locus in subsequent analyses as we could not rule out its role in other SUDs. In the cross-population meta-analysis, we identified 226 independent lead variants (135 loci) in the 1kg-AFR + EUR-like samples, 243 independent lead variants (113 loci) in the 1kg-EUR + AMR-like samples, and 144 independent lead variants (82 loci) in the 1kg-AFR + AMR + EUR-like samples (Supplemental Tables 5–7). Manhattan plots of each meta-analysis are displayed in Fig. 1 (only showing concordant variants). After merging loci identified in all meta-analyses if they had inter-loci distances <250 kb, there were 220 loci in total with 40 novel loci not previously reported as associated with any SUD related phenotype in the GWAS catalog [38] (Table 2). Among those 40 novel loci, 32 were associated with psychiatric traits such as insomnia, depression, and educational attainment, but the other eight were not associated with psychiatric traits in the GWAS catalog [38]. Thus, further studies are needed to confirm and understand these findings. LocusZoom plots of these 220 loci from the various meta-analyses are displayed in Supplemental Fig. 1 (only showing concordant variants). Notably, locus 158 was significant in both the 1kg-EUR-like and 1kg-AFR-like samples, but no other locus was significant in more than one population. Additionally, 152 loci were identified by multiple meta-analyses (Table 2), and many of these had different independent lead variants in different meta-analyses (Supplemental Fig. 1).
| Locus | Chr | Locus start | Locus end | Lead variant | BP | Effect Allele | Non-Effect Allele | Z | EUR.P-value | AFR.P-value | AMR.P-value | AFR.EUR.P-value | AMR.EUR.P-value | AFR.AMR.EUR.P.value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 1 | 28,634,681 | 29,169,593 | rs2236855 | 29,161,999 | A | C | 7.09 | 1.17E-09 | 6.20E-04 | 0.60 | 1.32E-12 | 2.32E-09 | 2.89E-12 |
| 2 | 1 | 32,168,381 | 32,221,913 | rs3766823 | 32,197,257 | A | G | 5.66 | 2.62E-07 | NA | 7.56E-04 | NA | 1.49E-08 | NA |
| 3 | 1 | 43,760,236 | 44,295,047 | rs951740 | 44,011,737 | A | G | 7.18 | 8.04E-12 | NA | 8.45E-03 | NA | 7.08E-13 | NA |
| 4 | 1 | 45,168,480 | 45,751,415 | rs35809187 | 45,412,276 | A | C | 5.68 | 3.54E-08 | NA | 0.09 | NA | 1.32E-08 | NA |
| 5 | 1 | 46,041,100 | 46,598,273 | rs61784801 | 46,179,022 | A | C | −5.98 | 3.49E-09 | 0.40 | 0.16 | 1.79E-07 | 2.30E-09 | 7.27E-08 |
| 6 | 1 | 49,463,925 | 50,591,851 | rs241501 | 49,765,438 | T | G | 5.95 | 2.75E-09 | 0.16 | NA | NA | NA | NA |
| 7 | 1 | 66,304,167 | 66,690,704 | rs7519259 | 66,434,743 | A | G | 9.65 | 4.72E-22 | NA | NA | NA | NA | NA |
| 8 | 1 | 71,489,252 | 71,920,314 | rs4650105 | 71,841,727 | A | G | 5.46 | 1.47E-07 | NA | 0.07 | NA | 4.89E-08 | NA |
| 9 | 1 | 72,749,726 | 72,954,641 | rs9425089 | 72,765,082 | A | C | −5.61 | 9.51E-07 | 0.01 | NA | 2.07E-08 | NA | NA |
| 10 | 1 | 73,275,828 | 74,131,749 | rs12031120 | 73,860,821 | A | G | −11.12 | 3.83E-24 | 8.61E-05 | 0.11 | 1.02E-28 | 5.78E-24 | 1.03E-27 |
| 11 | 1 | 79,278,918 | 79,488,107 | rs12754818 | 79,392,756 | A | C | −5.74 | 9.60E-09 | NA | NA | NA | NA | NA |
| 12 | 1 | 80,784,642 | 80,871,734 | rs67263181 | 80,858,910 | T | TA | −6.35 | 2.13E-10 | NA | NA | NA | NA | NA |
| 13 | 1 | 91,189,731 | 91,234,126 | rs1526480 | 91,209,986 | T | C | −6.86 | 9.43E-12 | 0.13 | 0.55 | 6.93E-12 | NA | NA |
| 14 | 1 | 97,490,949 | 97,605,851 | rs79313673 | 97,595,236 | T | C | 5.59 | 1.18E-05 | 1.59E-03 | 0.06 | 5.47E-08 | 5.05E-06 | 2.34E-08 |
| 15 | 1 | 97,894,619 | 97,944,059 | rs4950035 | 97,920,746 | T | C | 6.98 | 3.02E-12 | NA | NA | NA | NA | NA |
| 16 | 1 | 98,565,234 | 98,664,952 | rs61787321 | 98,632,436 | T | C | −5.45 | 4.96E-08 | NA | NA | NA | NA | NA |
| 17 | 1 | 165,086,665 | 165,126,202 | rs10753661 | 165,119,792 | A | G | −6.23 | 4.66E-10 | NA | 0.24 | NA | 4.72E-10 | NA |
| 18 | 1 | 171,809,976 | 171,906,903 | rs6425479 | 171,881,168 | A | G | −5.83 | 2.80E-08 | 0.01 | NA | 5.63E-09 | NA | NA |
| 19 | 1 | 173,773,928 | 175,003,074 | rs2142633 | 174,515,935 | A | G | 6.47 | 9.76E-11 | NA | 0.85 | NA | NA | NA |
| 20 | 1 | 197,312,162 | 197,821,769 | rs34451672 | 197,767,841 | CT | C | −5.52 | 3.41E-08 | NA | NA | NA | NA | NA |
| 21 | 1 | 227,217,412 | 227,512,190 | rs7556248 | 227,320,893 | T | C | 5.53 | 3.20E-08 | 0.16 | NA | 2.18E-07 | NA | NA |
| 22 | 1 | 236,868,733 | 236,905,585 | rs4659711 | 236,902,364 | T | C | 5.90 | 4.98E-09 | 0.08 | 0.70 | 3.55E-09 | NA | NA |
| 23 | 2 | 1,896,243 | 1,966,952 | rs10180442 | 1,958,049 | T | C | −5.84 | 5.22E-09 | NA | NA | NA | NA | NA |
| 24 | 2 | 4,824,073 | 4,909,777 | rs6725602 | 4,845,377 | T | C | −5.47 | 4.64E-08 | NA | NA | NA | NA | NA |
| 25 | 2 | 22,045,717 | 22,174,521 | rs368693555 | 22,129,730 | T | C | −5.86 | 1.20E-06 | 2.80E-03 | 0.42 | 4.77E-09 | 1.39E-06 | 5.99E-09 |
| 26 | 2 | 22,621,296 | 22,918,025 | rs1355175 | 22,761,803 | T | C | −6.07 | 1.29E-09 | NA | NA | NA | NA | NA |
| 27 | 2 | 23,541,309 | 23,582,960 | rs11674807 | 23,541,471 | T | C | −5.81 | 3.84E-08 | 0.05 | NA | 6.39E-09 | NA | NA |
| 28 | 2 | 27,498,734 | 27,752,871 | rs3739095 | 27,556,721 | A | G | 6.37 | 1.84E-10 | NA | NA | NA | NA | NA |
| 29 | 2 | 43,512,764 | 44,359,876 | rs75120545 | 44,271,496 | T | C | −6.38 | 1.75E-10 | NA | 0.94 | NA | 3.03E-10 | NA |
| 30 | 2 | 45,121,466 | 45,175,585 | rs472140 | 45,139,904 | T | C | 10.91 | 9.96E-28 | 0.27 | 0.14 | NA | 1.70E-27 | NA |
| 31 | 2 | 56,616,218 | 57,427,953 | rs72618684 | 56,828,238 | T | C | −6.36 | 8.48E-09 | NA | 4.97E-05 | NA | 2.08E-10 | NA |
| 32 | 2 | 57,942,987 | 58,505,679 | rs2678892 | 58,167,140 | A | G | 8.24 | 1.72E-16 | NA | NA | NA | NA | NA |
| 33 | 2 | 60,115,069 | 60,527,297 | rs359257 | 60,485,161 | A | C | −6.43 | 2.68E-09 | 0.02 | NA | 1.24E-10 | NA | NA |
| 34 | 2 | 61,429,568 | 61,843,183 | rs778766 | 61,786,848 | T | C | −5.59 | 2.31E-08 | NA | NA | NA | NA | NA |
| 35 | 2 | 62,710,273 | 62,780,325 | rs56044100 | 62,735,706 | A | G | −6.91 | 2.71E-09 | 3.97E-04 | NA | 4.75E-12 | NA | NA |
| 36 | 2 | 73,552,542 | 74,051,360 | rs1246094 | 73,707,335 | T | C | 6.18 | 6.52E-10 | NA | 0.52 | NA | 1.13E-09 | NA |
| 37 | 2 | 101,233,020 | 101,311,552 | rs4851350 | 101,234,362 | T | C | 5.98 | 9.64E-09 | NA | 0.04 | NA | 2.28E-09 | NA |
| 38 | 2 | 104,056,454 | 104,696,491 | rs4851030 | 104,160,261 | A | G | 7.71 | 1.51E-14 | NA | 0.04 | NA | 1.26E-14 | NA |
| 39 | 2 | 105,915,711 | 105,994,827 | rs4851758 | 105,952,417 | T | C | 6.26 | 2.63E-06 | 2.53E-05 | 0.24 | 3.97E-10 | 2.24E-06 | 4.08E-10 |
| 40 | 2 | 138,077,405 | 138,394,805 | rs2375518 | 138,270,224 | A | G | −5.82 | 3.87E-08 | NA | 0.03 | NA | 6.00E-09 | NA |
| 41 | 2 | 144,144,663 | 144,272,229 | rs4233566 | 144,272,229 | A | G | −7.00 | 2.64E-12 | NA | NA | NA | NA | NA |
| 42 | 2 | 145,341,259 | 145,568,183 | rs72991962 | 145,420,704 | A | C | −5.75 | 4.64E-07 | 8.50E-03 | 0.06 | 2.47E-08 | 1.44E-07 | 8.73E-09 |
| 43 | 2 | 147,828,502 | 147,981,914 | rs72859290 | 147,981,914 | A | G | 6.73 | 1.68E-11 | NA | NA | NA | NA | NA |
| 44 | 2 | 156,884,809 | 156,933,105 | rs72904197 | 156,887,825 | A | G | 6.02 | 1.78E-09 | NA | NA | NA | NA | NA |
| 45 | 2 | 159,310,846 | 159,539,677 | rs6437184 | 159,354,852 | T | C | −6.09 | 2.09E-06 | 1.04E-04 | NA | 1.13E-09 | NA | NA |
| 46 | 2 | 161,606,161 | 162,792,864 | rs57761252 | 161,865,998 | A | G | −7.07 | 3.44E-12 | 0.26 | 0.10 | 3.55E-09 | 1.51E-12 | 8.88E-09 |
| 47 | 2 | 164,733,530 | 164,928,899 | rs268257 | 164,906,685 | T | C | −5.72 | 1.04E-08 | NA | NA | NA | NA | NA |
| 48 | 2 | 178,565,913 | 178,608,411 | rs17400325 | 178,565,913 | T | C | 6.54 | 6.20E-11 | NA | NA | NA | NA | NA |
| 49 | 2 | 185,541,022 | 186,057,716 | rs1594166 | 185,787,586 | T | G | −6.34 | 9.97E-09 | 0.01 | NA | 2.36E-10 | NA | NA |
| 50 | 2 | 201,080,113 | 201,244,261 | rs3754797 | 201,242,293 | A | G | 7.48 | 1.02E-11 | 9.16E-03 | 0.04 | 1.97E-13 | 4.00E-12 | 7.55E-14 |
| 51 | 2 | 202,842,998 | 202,848,662 | rs55846056 | 202,848,214 | A | C | −5.61 | 2.08E-08 | 0.14 | 0.24 | 3.35E-08 | 4.35E-08 | 3.11E-08 |
| 52 | 2 | 208,219,840 | 208,300,941 | rs13392734 | 208,255,975 | A | C | −6.05 | 1.46E-09 | NA | 0.42 | NA | NA | NA |
| 53 | 2 | 226,336,692 | 226,400,416 | rs6436561 | 226,370,779 | A | G | 6.21 | 3.85E-08 | 0.03 | NA | 5.47E-10 | NA | NA |
| 54 | 3 | 16,729,269 | 17,119,238 | rs6800583 | 16,851,755 | A | G | 7.18 | 6.77E-13 | NA | NA | NA | NA | NA |
| 55 | 3 | 18,610,030 | 18,824,298 | rs9835977 | 18,793,340 | T | C | −7.01 | 2.38E-12 | NA | NA | NA | NA | NA |
| 56 | 3 | 24,902,655 | 24,923,743 | rs12495571 | 24,902,655 | T | C | −5.67 | 1.47E-08 | NA | 0.37 | NA | 1.48E-08 | NA |
| 57 | 3 | 48,412,246 | 49,940,078 | rs7625857 | 49,596,462 | T | C | −11.21 | 3.60E-29 | NA | 0.16 | NA | 5.71E-29 | NA |
| 58 | 3 | 53,840,207 | 53,858,541 | rs3774611 | 53,840,665 | A | G | −5.75 | 4.96E-08 | NA | 0.05 | NA | 8.92E-09 | NA |
| 59 | 3 | 55,936,417 | 56,414,387 | rs7641315 | 55,987,823 | A | G | −6.87 | 8.13E-12 | NA | 0.38 | NA | 6.46E-12 | NA |
| 60 | 3 | 59,777,008 | 59,786,032 | rs2629879 | 59,777,685 | A | G | −5.83 | 5.71E-09 | NA | 0.69 | NA | NA | NA |
| 61 | 3 | 81,236,994 | 81,985,189 | rs6804453 | 81,873,269 | T | C | −6.58 | 4.85E-11 | NA | 0.69 | NA | 1.27E-10 | NA |
| 62 | 3 | 84,841,679 | 84,951,716 | rs112987098 | 84,883,600 | CA | C | −7.31 | 2.71E-13 | NA | NA | NA | NA | NA |
| 63 | 3 | 85,481,557 | 86,209,081 | rs7644873 | 85,720,888 | T | C | −6.83 | 4.11E-10 | 0.01 | NA | 8.68E-12 | NA | NA |
| 64 | 3 | 98,489,915 | 98,770,189 | rs9811449 | 98,499,298 | T | G | 5.59 | 2.74E-08 | NA | 0.22 | NA | 2.32E-08 | NA |
| 65 | 3 | 104,647,183 | 104,805,201 | rs10804436 | 104,670,412 | A | C | −5.69 | 1.31E-08 | 0.18 | 0.37 | 6.56E-08 | 1.89E-08 | 6.56E-08 |
| 66 | 3 | 110,103,471 | 110,300,607 | rs6437918 | 110,225,767 | T | C | 6.39 | 4.83E-09 | 0.02 | 0.45 | 1.70E-10 | 7.07E-09 | 2.42E-10 |
| 67 | 3 | 114,135,165 | 114,233,728 | rs7630111 | 114,165,901 | A | C | 7.63 | 2.29E-14 | 0.04 | 0.13 | 3.84E-13 | 2.37E-14 | 1.61E-13 |
| 68 | 3 | 147,095,294 | 147,268,599 | rs2279829 | 147,106,319 | T | C | −6.04 | 1.54E-09 | NA | NA | NA | NA | NA |
| 69 | 3 | 157,829,953 | 158,534,111 | rs7629432 | 157,895,285 | T | C | 7.01 | 6.60E-09 | 4.74E-04 | 0.68 | 2.34E-12 | 1.34E-08 | 1.21E-11 |
| 70 | 3 | 161,703,030 | 161,801,528 | rs9868656 | 161,797,034 | T | C | 5.58 | 4.43E-07 | 0.03 | NA | 2.40E-08 | NA | NA |
| 71 | 4 | 28,247,071 | 28,572,856 | rs35236641 | 28,499,286 | A | AG | −5.60 | 2.11E-08 | NA | NA | NA | NA | NA |
| 72 | 4 | 42,036,458 | 42,185,424 | rs4861023 | 42,150,774 | A | G | −5.63 | 2.26E-07 | 0.05 | 0.18 | 3.48E-08 | 8.81E-08 | 1.78E-08 |
| 73 | 4 | 46,879,127 | 47,035,087 | rs10033762 | 46,907,959 | T | G | −7.46 | 9.82E-11 | 5.17E-03 | 0.04 | 1.15E-13 | 5.25E-11 | 8.88E-14 |
| 74 | 4 | 59,864,305 | 60,078,065 | rs78040411 | 59,919,626 | T | C | 5.85 | 4.82E-09 | NA | NA | NA | NA | NA |
| 75 | 4 | 62,007,737 | 62,098,937 | rs1394950 | 62,083,448 | T | C | −5.53 | 9.78E-07 | 8.47E-03 | NA | 3.20E-08 | NA | NA |
| 76 | 4 | 62,758,896 | 62,845,490 | rs734644 | 62,800,728 | T | C | −6.16 | 2.30E-09 | 0.15 | NA | 7.21E-10 | NA | NA |
| 77 | 4 | 85,485,247 | 85,965,865 | rs111669166 | 85,485,793 | A | G | −5.73 | 1.03E-08 | NA | NA | NA | NA | NA |
| 78 | 4 | 99,630,017 | 101,037,327 | rs1229984 | 100,239,319 | T | C | −9.48 | NA | 1.47E-05 | 2.57E-21 | NA | NA | NA |
| 79 | 4 | 102,702,364 | 103,387,161 | rs13107325 | 103,188,709 | T | C | −10.94 | 9.04E-27 | 0.41 | 2.18E-03 | 9.76E-22 | 7.70E-28 | 2.81E-23 |
| 80 | 4 | 112,208,748 | 112,503,872 | rs72678859 | 112,406,961 | T | C | −6.92 | 2.18E-11 | 0.11 | 0.73 | 4.59E-12 | NA | NA |
| 81 | 4 | 143,486,962 | 144,207,201 | rs2120162 | 143,547,293 | A | C | 7.21 | 2.48E-12 | NA | 0.02 | NA | 5.55E-13 | NA |
| 82 | 4 | 151,296,089 | 151,936,204 | rs12509782 | 151,605,828 | A | G | −6.02 | 1.70E-09 | NA | NA | NA | NA | NA |
| 83 | 4 | 152,238,328 | 152,724,537 | rs3792689 | 152,637,700 | T | C | 6.02 | 1.70E-09 | NA | NA | NA | NA | NA |
| 84 | 4 | 186,658,747 | 186,688,223 | rs56110010 | 186,658,747 | T | C | −5.92 | 1.16E-08 | NA | 0.04 | NA | 3.19E-09 | NA |
| 85 | 5 | 3,405,495 | 3,417,995 | rs10462767 | 3,417,930 | T | C | 6.28 | 6.84E-07 | 2.10E-03 | 1.32E-03 | 3.18E-09 | 7.87E-08 | 3.45E-10 |
| 86 | 5 | 30,783,801 | 30,845,465 | rs10940922 | 30,818,059 | T | C | 6.31 | 5.75E-10 | NA | 0.07 | NA | 2.85E-10 | NA |
| 87 | 5 | 43,054,747 | 43,193,052 | rs10941611 | 43,162,608 | T | C | −5.57 | 2.54E-08 | NA | NA | NA | NA | NA |
| 88 | 5 | 51,704,512 | 51,852,586 | rs11742575 | 51,822,955 | T | C | −6.44 | 2.82E-10 | 0.19 | NA | 1.23E-10 | NA | NA |
| 89 | 5 | 59,842,971 | 60,614,879 | rs7709444 | 60,095,726 | T | C | 7.30 | 5.44E-12 | 0.05 | 0.01 | 6.94E-13 | 9.64E-13 | 3.00E-13 |
| 90 | 5 | 61,397,967 | 61,959,339 | rs4700486 | 61,517,281 | A | G | −5.99 | 2.14E-09 | NA | NA | NA | NA | NA |
| 91 | 5 | 153,303,901 | 153,626,546 | rs2199626 | 153,507,230 | T | C | −6.73 | 4.68E-08 | 7.83E-04 | 0.03 | 3.16E-11 | 1.20E-08 | 1.72E-11 |
| 92 | 5 | 166,985,224 | 167,055,936 | rs1549212 | 166,996,722 | T | C | −6.92 | 9.65E-12 | 0.02 | NA | 4.49E-12 | NA | NA |
| 93 | 5 | 170,284,624 | 170,846,151 | rs4286697 | 170,454,681 | A | G | −7.06 | 1.43E-09 | 6.20E-04 | 0.59 | 1.63E-12 | 2.76E-09 | 3.47E-12 |
| 94 | 6 | 11,470,694 | 11,614,620 | rs821443 | 11,538,534 | T | C | −5.49 | 1.08E-06 | 0.02 | 0.67 | 4.05E-08 | 1.69E-06 | 6.41E-08 |
| 95 | 6 | 19,212,286 | 19,307,114 | rs1858513 | 19,228,938 | T | C | 5.74 | 9.58E-09 | NA | 0.79 | NA | 2.20E-08 | NA |
| 96 | 6 | 25,480,328 | 29,607,101 | rs71559024 | 27,352,750 | A | G | −8.42 | 3.92E-17 | 0.01 | NA | 1.04E-15 | NA | NA |
| 97 | 6 | 33,203,908 | 33,306,235 | rs464921 | 33,240,506 | A | G | −5.69 | 1.29E-08 | NA | 0.08 | NA | NA | NA |
| 98 | 6 | 37,609,457 | 37,632,322 | rs3846873 | 37,619,600 | A | G | 5.46 | 8.66E-06 | 9.20E-03 | 0.02 | 2.40E-07 | 1.41E-06 | 4.68E-08 |
| 99 | 6 | 47,802,504 | 47,885,971 | rs559028 | 47,809,908 | T | C | 6.57 | 5.18E-11 | NA | 0.82 | NA | NA | NA |
| 100 | 6 | 126,657,472 | 127,080,700 | rs9398809 | 126,789,144 | T | C | −5.73 | 2.95E-07 | 0.05 | 0.03 | 3.17E-08 | 9.01E-08 | 9.79E-09 |
| 101 | 6 | 143,141,800 | 143,197,874 | rs72995085 | 143,193,971 | T | C | 5.50 | 3.82E-08 | NA | NA | NA | NA | NA |
| 102 | 6 | 152,201,201 | 152,264,529 | rs6557171 | 152,234,593 | T | C | 5.52 | 3.33E-08 | NA | NA | NA | NA | NA |
| 103 | 6 | 154,215,240 | 154,390,607 | rs1799971 | 154,360,797 | A | G | 7.85 | 4.52E-14 | 0.12 | 1.80E-03 | 5.53E-13 | 4.03E-15 | 4.56E-14 |
| 104 | 6 | 162,931,993 | 163,209,495 | rs62430881 | 162,988,202 | A | C | −5.87 | 4.29E-08 | NA | 2.46E-03 | NA | 4.28E-09 | NA |
| 105 | 6 | 163,792,420 | 164,006,012 | rs1764021 | 163,805,682 | T | C | −6.67 | 2.64E-11 | NA | NA | NA | NA | NA |
| 106 | 6 | 169,799,110 | 170,248,792 | rs4236177 | 170,073,543 | T | C | 6.14 | 1.89E-09 | NA | 0.10 | NA | 8.05E-10 | NA |
| 107 | 7 | 75,607,155 | 75,910,452 | rs6467958 | 75,622,281 | T | C | −6.86 | 1.91E-11 | NA | 0.09 | NA | 6.93E-12 | NA |
| 108 | 7 | 97,264,194 | 97,310,892 | rs10275909 | 97,298,821 | A | G | −5.63 | 3.48E-08 | NA | 0.07 | NA | 1.79E-08 | NA |
| 109 | 7 | 109,928,395 | 110,028,559 | rs78971860 | 109,989,952 | T | C | −5.62 | 3.53E-06 | 4.83E-03 | NA | 1.88E-08 | NA | NA |
| 110 | 7 | 113,998,993 | 114,287,116 | rs2189010 | 114,119,430 | A | G | 7.88 | 3.19E-15 | NA | NA | NA | NA | NA |
| 111 | 7 | 114,940,159 | 115,025,708 | rs1155397 | 114,953,597 | A | G | 7.74 | 1.38E-14 | NA | 0.09 | NA | 1.02E-14 | NA |
| 112 | 7 | 117,497,811 | 117,593,308 | rs6965740 | 117,514,840 | T | G | −6.72 | 1.87E-11 | NA | NA | NA | NA | NA |
| 113 | 7 | 135,050,259 | 135,221,170 | rs11772832 | 135,073,047 | T | C | −8.36 | 1.33E-16 | NA | 0.26 | NA | 6.12E-17 | NA |
| 114 | 8 | 9,894,125 | 10,005,367 | rs7832431 | 9,947,661 | A | C | 5.96 | 2.22E-07 | 4.35E-03 | NA | 2.54E-09 | NA | NA |
| 115 | 8 | 12,662,159 | 12,694,291 | rs7812870 | 12,667,421 | T | C | −5.99 | 7.66E-07 | 6.08E-04 | NA | 2.06E-09 | NA | NA |
| 116 | 8 | 21,770,372 | 21,872,853 | rs2054711 | 21,818,060 | A | G | 6.97 | 3.96E-09 | 7.83E-04 | 0.18 | 5.66E-12 | 3.07E-09 | 3.28E-12 |
| 117 | 8 | 27,406,353 | 27,470,919 | rs11783093 | 27,425,349 | T | C | −7.82 | 5.51E-15 | NA | 0.17 | NA | NA | NA |
| 118 | 8 | 52,573,769 | 52,870,877 | rs2395827 | 52,786,700 | T | C | −5.52 | 4.54E-06 | 0.02 | 0.01 | 1.47E-07 | 1.04E-06 | 3.36E-08 |
| 119 | 8 | 57,221,079 | 57,437,154 | rs2246873 | 57,433,941 | A | G | −7.21 | 5.68E-13 | NA | NA | NA | NA | NA |
| 120 | 8 | 64,505,710 | 65,097,747 | rs7017705 | 64,982,843 | T | C | 6.77 | 3.89E-11 | NA | 0.05 | NA | 1.28E-11 | NA |
| 121 | 8 | 76,705,544 | 77,021,947 | rs830432 | 76,941,945 | A | G | −5.57 | 2.64E-06 | 6.41E-03 | 0.19 | 3.49E-08 | 1.84E-06 | 2.50E-08 |
| 122 | 8 | 77,588,548 | 77,695,732 | rs76132822 | 77,679,194 | T | C | 5.51 | 1.42E-07 | NA | 0.08 | NA | 3.57E-08 | NA |
| 123 | 8 | 87,139,254 | 87,241,715 | rs2954344 | 87,238,494 | T | G | −5.63 | 1.35E-05 | 3.00E-04 | NA | 1.80E-08 | NA | NA |
| 124 | 8 | 92,976,563 | 93,180,965 | rs9297901 | 93,036,795 | T | C | 7.03 | 2.95E-12 | 0.01 | NA | 2.10E-12 | NA | NA |
| 125 | 8 | 113,811,442 | 114,725,064 | rs6469450 | 114,179,599 | A | C | 7.18 | 6.80E-13 | NA | NA | NA | NA | NA |
| 126 | 8 | 142,021,655 | 142,031,024 | rs55793749 | 142,028,409 | A | G | −5.78 | 7.31E-09 | NA | NA | NA | NA | NA |
| 127 | 8 | 143,345,007 | 143,354,963 | rs7014279 | 143,352,528 | A | C | 5.62 | 1.96E-08 | NA | NA | NA | NA | NA |
| 128 | 9 | 16,089,196 | 16,102,759 | rs62552202 | 16,092,730 | A | C | 6.00 | 2.44E-08 | 0.03 | 0.42 | 2.00E-09 | NA | NA |
| 129 | 9 | 17,044,769 | 17,128,036 | rs425455 | 17,061,693 | T | C | −5.72 | 1.07E-08 | NA | NA | NA | NA | NA |
| 130 | 9 | 23,747,791 | 23,849,297 | rs9298850 | 23,765,511 | T | C | −6.30 | 1.13E-09 | NA | 0.07 | NA | 3.05E-10 | NA |
| 131 | 9 | 30,121,131 | 30,456,131 | rs12348434 | 30,247,380 | A | G | −6.47 | 5.12E-09 | 0.02 | 0.21 | 9.97E-11 | 4.99E-09 | 1.43E-10 |
| 132 | 9 | 81,331,339 | 81,508,309 | rs7029483 | 81,483,995 | A | G | −5.89 | 4.59E-08 | NA | 9.76E-04 | NA | 3.93E-09 | NA |
| 133 | 9 | 127,594,420 | 128,411,969 | rs4837011 | 127,923,014 | T | G | −9.73 | 2.27E-22 | 0.14 | 0.14 | 2.72E-21 | 3.75E-22 | 1.23E-20 |
| 134 | 9 | 134,267,022 | 134,269,789 | rs182293592 | 134,269,789 | T | G | 5.58 | 2.47E-08 | NA | NA | NA | NA | NA |
| 135 | 9 | 134,802,170 | 134,946,805 | rs9411334 | 134,896,887 | T | C | −5.78 | 1.58E-08 | NA | 0.07 | NA | 7.49E-09 | NA |
| 136 | 9 | 136,876,021 | 136,973,826 | rs467387 | 136,907,005 | A | G | 5.96 | 3.65E-08 | 7.54E-03 | NA | 2.51E-09 | NA | NA |
| 137 | 10 | 10,559,149 | 10,700,036 | rs7924306 | 10,573,209 | T | G | 5.63 | 6.13E-07 | 0.02 | NA | 1.78E-08 | NA | NA |
| 138 | 10 | 64,874,754 | 65,400,271 | rs7075302 | 65,381,740 | A | G | 6.07 | 1.57E-07 | 7.56E-03 | 0.14 | 1.30E-09 | NA | NA |
| 139 | 10 | 75,430,874 | 75,661,162 | rs12775558 | 75,570,868 | A | G | 6.41 | 8.35E-09 | 0.01 | NA | 1.47E-10 | NA | NA |
| 140 | 10 | 80,094,714 | 80,159,503 | rs938244 | 80,113,953 | T | C | −5.67 | 1.97E-07 | 0.08 | 0.06 | 3.41E-08 | 8.44E-08 | 1.46E-08 |
| 141 | 10 | 84,735,682 | 84,968,882 | rs10885703 | 84,843,021 | T | C | 5.63 | 1.80E-08 | NA | 0.41 | NA | 2.49E-08 | NA |
| 142 | 10 | 102,912,264 | 105,165,256 | rs1046411 | 104,837,816 | A | G | 6.36 | 1.03E-07 | 4.95E-03 | 9.72E-03 | 9.63E-10 | 2.12E-08 | 1.96E-10 |
| 143 | 10 | 110,461,674 | 110,758,956 | rs7074871 | 110,507,806 | A | G | −9.04 | 8.20E-14 | 9.85E-06 | 3.90E-03 | 4.19E-19 | 1.02E-14 | 1.53E-19 |
| 144 | 10 | 118,902,004 | 118,919,575 | rs73385301 | 118,909,289 | A | G | 6.24 | 1.43E-09 | NA | 0.09 | NA | 4.37E-10 | NA |
| 145 | 11 | 24,932,312 | 25,098,152 | rs12278803 | 24,941,827 | A | G | 5.66 | 1.55E-08 | NA | NA | NA | NA | NA |
| 146 | 11 | 27,646,247 | 27,736,207 | rs1808124 | 27,686,904 | T | C | 5.49 | 4.01E-08 | NA | NA | NA | NA | NA |
| 147 | 11 | 27,996,573 | 28,709,434 | rs10835368 | 28,643,299 | T | C | −6.22 | 9.78E-10 | NA | 0.12 | NA | 5.15E-10 | NA |
| 148 | 11 | 38,401,216 | 38,998,930 | rs2007212 | 38,490,584 | T | C | −6.48 | 9.35E-11 | NA | 0.86 | NA | NA | NA |
| 149 | 11 | 46,336,995 | 46,751,495 | rs7476 | 46,342,834 | A | C | −5.86 | 4.61E-09 | 0.17 | NA | 5.99E-08 | NA | NA |
| 150 | 11 | 47,372,377 | 47,946,836 | rs7107356 | 47,676,170 | A | G | −6.70 | 6.03E-10 | 0.02 | 0.40 | 2.10E-11 | 8.80E-10 | 2.93E-11 |
| 151 | 11 | 57,385,856 | 57,681,828 | rs12785533 | 57,486,651 | A | G | 7.31 | 2.61E-13 | NA | NA | NA | NA | NA |
| 152 | 11 | 63,869,596 | 64,218,143 | rs113421210 | 64,133,658 | T | G | 7.58 | 5.77E-13 | 0.25 | 0.01 | 3.89E-12 | 3.47E-14 | 3.80E-12 |
| 153 | 11 | 73,305,859 | 73,312,252 | rs4433590 | 73,310,668 | A | G | −5.78 | 7.34E-09 | NA | 0.25 | NA | 1.77E-08 | NA |
| 154 | 11 | 88,470,041 | 89,055,293 | rs681820 | 88,615,547 | A | G | −5.99 | 2.08E-09 | NA | 0.72 | NA | NA | NA |
| 155 | 11 | 95,440,480 | 95,681,028 | rs1150361 | 95,633,805 | A | G | −7.38 | 1.19E-11 | 0.01 | 5.02E-03 | 8.74E-13 | 1.36E-12 | 1.54E-13 |
| 156 | 11 | 112,826,311 | 114,008,053 | rs61902812 | 113,374,420 | A | C | −11.81 | 1.40E-29 | 4.44E-03 | 2.26E-03 | 3.66E-32 | 8.50E-30 | 3.40E-32 |
| 157 | 11 | 116,073,168 | 116,103,543 | rs10891932 | 116,103,543 | A | G | 7.03 | 1.18E-10 | 3.94E-03 | 3.99E-03 | 1.63E-11 | 1.30E-11 | 2.15E-12 |
| 158 | 11 | 121,495,141 | 121,662,150 | rs1944678 | 121,646,580 | A | C | 9.44 | 5.40E-13 | 9.49E-09 | 0.02 | 5.25E-20 | 4.57E-14 | 3.69E-21 |
| 159 | 11 | 126,898,083 | 126,998,347 | rs12362296 | 126,983,715 | A | G | −6.44 | 1.17E-10 | NA | NA | NA | NA | NA |
| 160 | 11 | 132,038,207 | 132,321,312 | rs11223043 | 132,254,814 | A | G | −5.94 | 7.15E-07 | 5.99E-03 | 0.05 | 4.64E-09 | 2.86E-07 | 2.90E-09 |
| 161 | 11 | 133,303,819 | 133,370,476 | rs6590712 | 133,349,706 | A | C | 5.68 | 2.50E-06 | 9.08E-04 | 0.34 | 2.28E-08 | 2.54E-06 | 1.33E-08 |
| 162 | 12 | 2,285,731 | 2,402,246 | rs11062148 | 2,298,085 | T | C | 5.51 | 3.57E-08 | NA | 0.46 | NA | NA | NA |
| 163 | 12 | 51,780,136 | 51,919,819 | rs10783447 | 51,884,821 | A | G | −8.29 | 3.13E-16 | NA | 0.02 | NA | 1.10E-16 | NA |
| 164 | 12 | 53,730,164 | 54,028,059 | rs61928096 | 53,780,633 | A | G | 5.46 | 6.75E-07 | 0.05 | 0.36 | 7.73E-08 | 5.23E-07 | 4.70E-08 |
| 165 | 12 | 81,469,429 | 81,673,764 | rs35166008 | 81,571,248 | A | G | −6.08 | 1.22E-09 | NA | NA | NA | NA | NA |
| 166 | 12 | 83,863,748 | 84,165,561 | rs1599264 | 84,122,022 | A | G | −5.49 | 4.07E-08 | NA | NA | NA | NA | NA |
| 167 | 12 | 110,501,407 | 111,281,108 | rs7971825 | 111,061,652 | A | G | 5.59 | 2.22E-08 | NA | 0.73 | NA | NA | NA |
| 168 | 12 | 120,879,294 | 121,185,145 | rs12829722 | 121,155,622 | T | C | 6.09 | 2.53E-06 | 4.64E-04 | 0.02 | 3.44E-09 | 6.98E-07 | 1.17E-09 |
| 169 | 12 | 122,154,617 | 122,203,680 | rs28651018 | 122,186,268 | T | C | −5.89 | 3.93E-09 | 7.92E-03 | NA | NA | NA | NA |
| 170 | 13 | 55,603,644 | 56,250,005 | rs145720647 | 55,708,612 | A | G | 8.19 | 3.18E-12 | 2.45E-04 | 0.02 | 3.69E-16 | 7.32E-13 | 2.64E-16 |
| 171 | 13 | 79,855,297 | 80,141,911 | rs9318638 | 79,986,153 | T | C | −6.15 | 2.18E-07 | 1.65E-03 | NA | 7.65E-10 | NA | NA |
| 172 | 13 | 96,589,737 | 97,029,792 | rs4771935 | 96,996,902 | A | C | −7.28 | 4.89E-10 | 1.85E-03 | 0.06 | 3.38E-13 | 2.22E-10 | 4.52E-13 |
| 173 | 14 | 41,952,029 | 42,183,025 | rs12431444 | 42,068,689 | T | C | 6.21 | 3.63E-09 | NA | 3.45E-03 | NA | 5.19E-10 | NA |
| 174 | 14 | 47,243,109 | 47,409,360 | rs12889080 | 47,291,881 | T | C | 5.50 | 4.52E-07 | NA | 1.46E-03 | NA | 3.76E-08 | NA |
| 175 | 14 | 55,569,578 | 55,936,577 | rs17743595 | 55,901,926 | A | G | 5.50 | 2.82E-07 | 0.04 | 0.51 | 3.81E-08 | 4.06E-07 | 5.93E-08 |
| 176 | 14 | 57,338,977 | 57,381,487 | rs864279 | 57,349,958 | T | C | 7.11 | 1.20E-12 | NA | 0.31 | NA | NA | NA |
| 177 | 14 | 58,664,909 | 58,856,709 | rs6573198 | 58,759,762 | A | G | −6.74 | 5.06E-11 | 0.08 | 0.23 | 1.70E-11 | 5.29E-11 | 1.59E-11 |
| 178 | 14 | 79,453,862 | 79,614,043 | rs10143998 | 79,525,531 | T | C | −6.70 | 6.10E-08 | 1.18E-04 | 0.29 | 2.16E-11 | 5.85E-08 | 2.34E-11 |
| 179 | 14 | 99,733,954 | 99,750,520 | rs1405238 | 99,733,954 | T | C | 6.14 | 8.04E-10 | NA | NA | NA | NA | NA |
| 180 | 14 | 104,021,141 | 104,363,528 | rs6576006 | 104,322,394 | A | C | 6.81 | 1.68E-11 | 5.30E-03 | NA | 9.79E-12 | NA | NA |
| 181 | 15 | 47,613,403 | 47,971,793 | rs175979 | 47,674,812 | A | G | 9.29 | 5.62E-18 | 1.34E-03 | NA | 1.59E-20 | NA | NA |
| 182 | 15 | 82,443,939 | 82,537,354 | rs7162177 | 82,474,069 | T | C | 6.34 | 2.26E-10 | NA | NA | NA | NA | NA |
| 183 | 15 | 83,413,041 | 83,977,166 | rs35563832 | 83,904,049 | T | G | −7.52 | 5.45E-14 | NA | 0.75 | NA | 4.25E-13 | NA |
| 184 | 15 | 91,416,550 | 91,429,042 | rs4702 | 91,426,560 | A | G | −6.26 | 3.76E-10 | NA | NA | NA | NA | NA |
| 185 | 16 | 13,580,639 | 13,763,942 | rs12925872 | 13,752,356 | T | C | 5.67 | 3.79E-06 | 4.91E-03 | 0.03 | 3.66E-08 | 1.25E-06 | 1.47E-08 |
| 186 | 16 | 19,246,136 | 19,279,464 | rs12921478 | 19,272,075 | A | C | −6.75 | 1.45E-11 | NA | 0.53 | NA | NA | NA |
| 187 | 16 | 23,758,347 | 23,901,376 | rs194795 | 23,843,121 | T | G | −5.80 | 6.55E-09 | NA | NA | NA | NA | NA |
| 188 | 16 | 29,924,422 | 30,120,442 | rs8060511 | 30,101,596 | A | C | −7.71 | 7.55E-14 | 0.05 | 0.35 | 1.26E-14 | 4.69E-14 | 1.96E-14 |
| 189 | 16 | 53,797,908 | 53,848,561 | rs56137030 | 53,825,905 | A | G | −10.12 | 8.17E-23 | 0.01 | 0.12 | 4.35E-24 | 1.14E-22 | 2.39E-23 |
| 190 | 16 | 61,583,172 | 61,650,572 | rs9635513 | 61,631,362 | T | C | 5.50 | 1.52E-07 | 0.10 | NA | 3.72E-08 | NA | NA |
| 191 | 16 | 72,052,034 | 72,249,332 | rs8047881 | 72,193,906 | A | G | −5.54 | 5.41E-07 | 0.04 | NA | 3.12E-08 | NA | NA |
| 192 | 16 | 73,594,142 | 73,606,553 | rs825704 | 73,603,649 | T | C | 5.66 | 3.24E-07 | 0.02 | 0.21 | 1.70E-08 | 1.82E-07 | 1.48E-08 |
| 193 | 16 | 76,419,114 | 76,519,827 | rs1506812 | 76,507,956 | A | C | −6.32 | 3.09E-09 | 0.04 | NA | 2.62E-10 | NA | NA |
| 194 | 16 | 87,430,239 | 87,518,815 | rs3748400 | 87,445,839 | T | C | −5.54 | 2.41E-07 | NA | 0.02 | NA | 3.04E-08 | NA |
| 195 | 17 | 7,802,968 | 7,924,748 | rs12943962 | 7,818,004 | T | G | −5.55 | 9.54E-07 | 0.02 | 0.72 | 2.88E-08 | 1.63E-06 | 2.89E-08 |
| 196 | 17 | 18,040,690 | 18,128,474 | rs854784 | 18,040,690 | T | C | 6.34 | 1.17E-08 | 0.01 | 0.01 | 2.34E-10 | NA | NA |
| 197 | 17 | 29,389,026 | 29,735,829 | rs2525570 | 29,681,245 | A | G | −6.42 | 1.71E-09 | 0.03 | NA | 1.33E-10 | NA | NA |
| 198 | 17 | 44,040,626 | 44,056,097 | rs3785883 | 44,054,433 | A | G | 5.63 | 3.28E-08 | 0.13 | 0.37 | 2.92E-08 | 2.76E-08 | 1.80E-08 |
| 199 | 17 | 48,200,834 | 48,234,411 | rs847680 | 48,224,075 | T | C | −5.55 | 2.35E-07 | 0.13 | NA | 2.91E-08 | NA | NA |
| 200 | 17 | 55,655,932 | 55,675,058 | rs72833102 | 55,663,552 | A | G | 7.19 | 1.71E-12 | NA | 0.13 | NA | 6.62E-13 | NA |
| 201 | 17 | 65,822,573 | 66,098,979 | rs62084675 | 66,027,025 | A | G | 5.86 | 4.62E-09 | NA | 1.82E-03 | NA | NA | NA |
| 202 | 18 | 35,125,113 | 35,204,858 | rs17570353 | 35,154,734 | A | G | 6.33 | 4.38E-09 | 0.05 | NA | 2.52E-10 | NA | NA |
| 203 | 18 | 39,253,947 | 39,322,811 | rs112226573 | 39,305,644 | T | TCTC | 5.49 | 4.04E-08 | NA | NA | NA | NA | NA |
| 204 | 18 | 50,555,225 | 51,054,765 | rs35916734 | 50,762,061 | A | ATT | 6.27 | 3.59E-10 | NA | NA | NA | NA | NA |
| 205 | 18 | 52,369,516 | 52,523,477 | rs60512641 | 52,523,477 | A | G | 7.19 | 2.77E-12 | 3.68E-03 | NA | 6.51E-13 | NA | NA |
| 206 | 18 | 53,195,249 | 53,463,661 | rs12458015 | 53,305,735 | T | C | −6.06 | 3.77E-09 | 0.11 | NA | 1.40E-09 | NA | NA |
| 207 | 18 | 77,551,586 | 77,580,712 | rs11664298 | 77,578,986 | A | G | 5.64 | 1.69E-08 | NA | NA | NA | NA | NA |
| 208 | 19 | 612,234 | 637,909 | rs41542013 | 623,728 | T | C | −5.69 | 6.40E-06 | 2.07E-03 | 0.07 | 4.37E-08 | 1.62E-06 | 1.27E-08 |
| 209 | 19 | 5,033,752 | 5,084,579 | rs16992771 | 5,046,070 | T | C | −6.32 | 2.57E-10 | NA | 0.94 | NA | 9.20E-10 | NA |
| 210 | 19 | 10,744,807 | 10,853,296 | rs10405617 | 10,752,968 | A | G | −6.22 | 5.13E-10 | 0.26 | NA | 9.54E-10 | NA | NA |
| 211 | 19 | 18,519,997 | 18,637,194 | rs2278238 | 18,576,484 | T | C | −6.22 | 2.28E-06 | 3.28E-05 | 0.35 | 5.00E-10 | 2.19E-06 | 5.64E-10 |
| 212 | 19 | 19,355,537 | 19,643,907 | rs12463074 | 19,445,869 | A | G | 5.49 | 1.67E-07 | NA | 0.06 | NA | 4.12E-08 | NA |
| 213 | 19 | 49,168,942 | 49,254,955 | rs2548458 | 49,209,325 | T | C | 7.52 | 1.41E-12 | 0.02 | 0.23 | 5.59E-14 | 1.52E-12 | 1.37E-13 |
| 214 | 20 | 31,093,514 | 31,187,504 | rs911527 | 31,165,105 | T | C | 5.78 | 7.41E-09 | NA | NA | NA | NA | NA |
| 215 | 20 | 41,956,451 | 42,016,520 | rs6030812 | 42,001,011 | A | G | −6.81 | 2.99E-10 | 0.04 | NA | 9.54E-12 | NA | NA |
| 216 | 20 | 48,435,242 | 48,645,999 | rs7272308 | 48,583,726 | A | G | 6.28 | 5.61E-09 | NA | 3.88E-03 | NA | 3.45E-10 | NA |
| 217 | 21 | 40,512,918 | 40,718,354 | rs114399588 | 40,653,958 | A | G | 6.22 | 4.07E-07 | 3.69E-03 | 0.49 | 5.10E-10 | NA | NA |
| 218 | 21 | 46,485,406 | 46,551,702 | rs2150438 | 46,485,941 | A | G | 6.52 | 5.04E-10 | 6.45E-03 | NA | 7.28E-11 | NA | NA |
| 219 | 22 | 42,363,812 | 42,690,311 | rs9306356 | 42,567,090 | T | C | 5.85 | 1.02E-08 | NA | 0.26 | NA | 4.84E-09 | NA |
| 220 | 22 | 49,324,829 | 49,355,688 | rs5770456 | 49,349,639 | T | G | −5.78 | 7.60E-09 | NA | 0.30 | NA | NA | NA |
By design, we only included variants with concordant effects; however, some concordant GWS variants also showed certain degree of associations (e.g., P-value <0.05) across all SUDs. For example, there were 2999 (1kg-EUR-like), 8 (1kg-AFR-like), 11 (1kg-AFR + EUR-like), and 146 (1kg-AMR + EUR-like) GWS variants with P-values <0.05 in all SUD GWAS (Supplemental Table 2, 3, 5, and 6). No GWS variants had P-values <0.05 in all SUD GWAS in the 1kg-AMR-like and 1kg-AFR + AMR + EUR-like samples.
Gene-based analysis
We used MAGMA [33] and identified 494 significant genes in the 1kg-EUR-like sample, one significant gene in the 1kg-AFR-like sample, four significant genes in the 1kg-AMR-like sample, 552 significant genes in the 1kg-AFR + EUR-like samples, 492 significant genes in the 1kg-AMR + EUR-like samples, and 526 significant genes in the 1kg-AFR + AMR + EUR-like samples (Supplement tables 8–13). No gene was significant in more than one population. In total, 637 unique genes were identified by gene-based analysis. Manhattan plots of gene-based analyses are shown in Supplemental Fig. 2.
Mapping significant variants to genes
We used positional mapping, eQTL mapping, and chromatin interaction mapping and identified 711 genes in the 1kg-EUR-like sample, seven genes in the 1kg-AFR-like sample, six genes in the 1kg-AMR-like sample, 495 genes in the 1kg-AFR + EUR-like samples, 482 in 1kg-AMR + EUR-like samples, and 373 genes in the 1kg-ARF + AMR + EUR-like samples (Supplemental Tables 14–18). In total, we identified 837 unique genes.
Gene prioritization
We used a combination of gene-based analysis and gene mapping and identified a total of 1,044 genes. Among them, 785 genes were prioritized, which included 702 protein coding genes, 35 non-coding RNAs, 31 antisense genes, and 12 pseudogenes (Supplemental Table 19). Three genes were prioritized in both the 1kg-EUR-like and 1kg-AMR-like samples (ADH1C, ADH5, and RP11-696N14.1), and no other genes were prioritized in more than one population.
Brain dissection and cell type enrichment analyses
Of the 785 prioritized genes, they were highly expressed in 48 dissections with one from amygdala, two from basal nuclei, 26 from cortex, one from epithalamus, five from hippocampus, four from hypothalamus, one from midbrain, two from paleocortex, and six from thalamus (Supplemental Table 20). Prioritized genes with low expression levels were observed in 17 dissections with one from basal forebrain, two from basal nuclei, three from cerebellum, two from hippocampus, two from midbrain, three from myelencephalon, one from perirhinal cortex, two from pons, and one from spinal cord (Supplemental Table 21). Prioritized genes were highly expressed in 251 of the 461 cell types with all but four in neuronal cells (Supplemental Table 22). Most of these cell types were from amygdala, basal forebrain, cerebral cortex, hippocampus, hypothalamus, and thalamus, which is in agreement with what was observed in the dissection enrichment analysis. Prioritized genes with low expression levels were observed in three cell types (astrocyte, microglia, and vascular) (Supplemental Table 23).
SNP-heritability estimation and calculating SNP-heritability explained by concordant variants for each SUD
There were 799,022 concordant variants (representing 11.76, 15.17, 11.26, and 9.97% of the total number of variants in PAU, OUD, CUD, and TUD, respectively) in the 1kg-EUR-like sample. Using these concordant variants, we estimated a SNP-heritability of SUD (using LD score regression [41]) of 0.10 (SE = 0.006). These concordant variants explained 77.96%, 95.95%, 83.88%, and 56.01% of SNP-heritability of PAU, OUD, CUD and PTU, respectively in the 1kg-EUR-like sample (Supplemental Table 24).
Genetic correlations
We identified 567 significant correlations in our cross SUD meta-analysis in the 1kg-EUR-like sample (Supplemental Table 25). Among those correlations, 24 were related substance use, 35 were related to psychiatric diseases, 46 were related to behavior, and six were related to life events.
Polygenic score analyses
In the 1kg-AFR-like sample, there were 69,159 individuals in the AOU dataset (12,353 cases) and 1829 in the IB dataset (912 cases). In the 1kg-AMR-like sample, the AOU dataset included 62,489 individuals (5925 cases). In the 1kg-EUR-like sample, the AOU dataset comprised 209,615 individuals (20,910 cases) and the IB dataset included 6375 individuals (2801 cases) (Supplemental Table 26). PGS were significant in all analyses except in the 1kg-AFR-like IB samples (Fig. 2, Supplemental Table 26). In both the 1kg-AMR-like and 1kg-EUR-like samples, individuals in the top 10% of genetic risk were approximately twice as likely to have SUDs as compared to those in the remaining 90%, across both the AOU and IB datasets. This pattern was consistent in analyses combining males and females, as well as in sex-stratified analyses, with odds ratios (OR) ranging from 1.95 to 2.87. PGS were slightly higher in females than in males, but the 95% CIs overlapped. In the meta-analyses, all PGS were significant, and results were similar to AOU.
Drug repurposing
By querying DGIdb [50], we identified 260 FDA approved drugs with ATC code N targeting 51 genes (Supplemental Table 27). Seven drugs met the criteria for repurposing: Topiramate in ATC N03AX (treating epilepsy, targeting GABRA4); Aripiprazole and Cariprazine in ATC N05AX (treating schizophrenia, bipolar disorder, and major depressive disorder, targeting DRD2); Desipramine, Imipramine, and Nortriptyline in ATC N06AA (treating depression and anxiety, targeting BDNF); and Methylphenidate in ATC N06BA (treating ADHD, targeting BDNF). In the large-scale real-world data from the Clinformatics® database, users of these medications exhibited lower hazard ratios for developing SUDs compared to users of comparator drugs (Supplemental Table 28). For example, after adjusting for comorbidities, the hazard ratios were 0.44 for users of Topiramate (95% CI: 0.42–0.47), 0.88 for users of Aripiprazole or Cariprazine (95% CI: 0.78–0.88), 0.89 for users of Desipramine, Imipramine, or Nortriptyline (95% CI: 0.84–0.94); and 0.84 for users of Methylphenidate (95% CI: 0.78–0.91). These findings suggest that these seven drugs may have potential for repurposing in the treatment of SUDs.
Discussion
In this study, we identified 220 SUD-shared loci in multiple populations, with the majority of variants detected in the 1kg-EUR-like population, likely due to its substantially larger sample size. Of these loci, 40 had not been previously reported as SUD related in the GWAS catalog [38] and thus represent novel associations. All identified loci showed consistent directions of effect across individual SUDs, supporting their role as shared genetic risk factors. Through gene-based analyses, gene mapping, and gene prioritization, we identified 785 SUD-shared genes. These genes are highly expressed in amygdala, cortex, hippocampus, hypothalamus, and thalamus, primarily in neuronal cells. Identified concordant variants explained 56–96% of SNP-heritability of each SUD in the 1kg-EUR-like sample. In the 1kg-EUR-like and 1kg-AMR-like samples, the top 10% of individuals with the highest PGS were approximately twice as likely to have an SUD as compared to the remaining 90% in their groups. Lastly, we identified seven FDA approved drugs to treat other diseases that could potentially be repurposed to treat SUDs.
Identifying genes shared across different SUDs is challenging due to the heterogeneity of individual SUD diagnoses, which are each based on multiple criteria. Combining SUDs as one phenotype to identify SUD-shared gene make it even more heterogenous. Consequently, some findings may be SUD-specific but could be mistakenly identified as SUD-shared due to their small P-values. Since SUD-specific variants may have discordant effects in other SUDs, only considering concordant variants removed most of them. Additionally, we did not use LD score regression to estimate the common factor, and thereby avoided findings due to model misspecifications or LD pattern mismatches and increased the statistical power. We validated our findings through the out-of-sample PGS analyses using two independent samples. The PGS had high predictive power in the 1kg-EUR-like and 1kg-AMR-like samples, demonstrating that the concordant variants included in calculating PGS were likely SUD-shared. For comparison purposes, we also performed cross SUD GWAS by using genomic SEM in 1kg-EUR-like sample (data not shown). Genomic SEM identified 161 loci and 186 lead variants, fewer than those identified by using meta-analysis but only considering concordant variants (184 loci and 428 lead variants). For those top 10% having the highest PGS calculated using genomic SEM results, the OR was 1.77 (95% CI: 1.47–2.12) in the IB 1kg-EUR-like sample, which was smaller than the OR estimated by using concordant variants from meta-analysis (OR = 2.0, 95% CI: 1.67–2.40).
We prioritized 785 genes. The large numbers of genes are expected as SUDs are highly polygenic [25, 26, 60–65]. All prioritized genes were expressed in brain tissue and associated with psychiatric or brain related traits, making them promising targets for functional validation studies to investigate their putative roles in SUD. When calculating genetic correlations using LD score regression, all variants included are assumed to have contributions to the trait [18]; therefore, by retaining concordant variants, most irrelevant variants were excluded from LD score regression and thus increased the power and accuracy of genetic correlation analyses. Some of the identified genetic correlations are worth noting. For instance, while traits such as going to the pub or social club were positively correlated with SUDs (genetic correlation (rG) = 0.67, P-value = 3.30E-41), traits such as going to a sports club or the gym, or attending group activities such as adult education classes and religious groups were negatively correlated with SUDs (rG < = −0.40, P-value <3.65E-09). This suggests that the latter leisure/social activities can be used to reduce the incidence of SUDs. In addition, maternal smoking around birth was highly positively correlated with SUDs (rG = 0.91, P-value = 2.28E-125), indicating a significant maternal genetic contribution to the future development of SUDs in offspring.
Koob and Volkow proposed a heuristic framework of SUDs based on animal studies and human brain imaging data [3]. The framework proposed three brain regions related to three stages of SUDs: prefrontal cortex (preoccupation/anticipation stage), basal ganglia (binge/intoxication stage), and extended amygdala (withdrawal/negative affect stage) [3]. We found that our prioritized genes were highly expressed in prefrontal cortex, in support of the proposed framework. However, we also observed that among all 34 dissections from cortex, prioritized genes were highly expressed in 26 of them. This suggests that most parts of the cortex may be related to SUDs, not just the prefrontal cortex. We found prioritized genes were highly expressed in caudate and putamen, but not other basal nuclei dissections. Prioritized genes were highly expressed in one amygdala dissection but not in the extended amygdala. Finally, while prioritized genes were not highly expressed in some basal nuclei dissections and extended amygdala, they were not expressed at particularly low levels in these regions either. Therefore, they were not necessarily unrelated to SUDs but may play less important roles in SUDs than other regions genetically (but may still play important roles via non-genetic mechanisms). Collectively, our findings indicate that more brain regions may be related to SUDs than previous studies have suggested [3]. We found that prioritized genes were highly expressed mostly in neuronal cells, indicating that the expression of SUD related genes is also cell type specific. To the best of our knowledge, this is the first comprehensive study of SUD-shared genes in different cell types in human brain. Our work sheds light on the etiologies of SUDs and further research is needed to confirm our findings.
In the 1kg-EUR-like and 1kg-AMR-like samples, the ORs for developing SUDs when comparing the top 10% individuals with the highest PGS to the remaining samples were approximately 2 or greater. These are the first SUD PGS that meet the suggested threshold for incorporation into clinical settings to provide additional information for the risk evaluation [66, 67]. In the 1kg-AFR-like sample, although PGS were statistically significant in both the AOU dataset and meta-analyses, the observed ORs ranged between 1.16 and 1.19, indicating that substantially larger sample sizes are needed to improve power. It is worth noting that the size of the 1kg-AMR-like sample was much smaller than that of the 1kg-AFR-like sample, but their PGS had much higher predictive power. One possible explanation for this result is that the 1kg-AMR-like sample is genetically closer to the 1kg-EUR-like sample, and thus using the 1kg-EUR-like sample GWAS significantly increased the statistical power. PGS meta-analysis results were similar to those in the AOU dataset. This is expected as the AOU dataset had much larger sample sizes than the IB dataset. In all analyses, PGS had slightly better predictive power in females than in males. Since PGS were calculated using autosomal variants only; we expected the PGS to perform similarly in males and females. The slight difference in our analyses and the different prevalence could be due to variants on the X and Y chromosomes, but more importantly, interactions between genetic and environmental factors. Adjusting PGS in epidemiological studies could help identify non-genetic factors that reduce the incidence of SUDs in females and thus facilitate the development of novel prevention strategies.
In the 1kg-EUR-like and 1kg-AFR-like samples, for those at high genetic risk, early interventions could be applied to reduce the incidence of SUDs. Additionally, PGS can be measured at any time, even before the initiation of risky substance use or risky behavior to further improve the effectiveness of prevention/intervention programs. PGS can also be used to stratify patients based on their risk levels to provide personalized treatments and improve the probability of recovery. That said, we want to caution readers that the predictive power of PGS is bounded by the estimated heritability of SUDs (~50%). Therefore, only using PGS will generate unreliable risk estimates as environmental factors are not considered. To reliably predict the risk for SUDs, both PGS and environmental factors, as well as their interactions should be carefully considered.
We used a large-scale real-world dataset and a robust active comparator design to identify repurposable drugs. For those seven drugs identified, previous studies have demonstrated their potential in treating just one or a few specific SUDs [51–58]—for example, Topiramate has been shown to reduce alcohol consumption [51, 68]. In contrast, our results suggest that these drugs may have broader therapeutic potential across all SUDs, as they target genes shared among multiple SUDs. This is especially crucial for affected individuals suffering from comorbid SUDs. It is worth noting that with the exception of Topiramate, the other six drugs target BDNF and DRD2-- genes that are also targeted by FDA-approved medications to treat substance-specific SUDs such as Bupropion (tobacco), Disulfiram (alcohol), and Methadone (opioids). This may facilitate their approval process, and these drugs should be prioritized for testing their efficacy in clinical trials.
Our study has several limitations. First, while considering concordant variants makes our findings more easily interpretable, it will miss those SUD-shared variants that are not present in a study due to a variety of reasons (e.g. not passing QC). Second, the sample sizes of both the 1kg-AFR-like and 1kg-AMR-like discovery and target datasets are small to moderate. This resulted in limited statistical power and thus genetic variants with small effects may not be detected. Consequently, the majority of our findings were from the 1kg-EUR-like sample and some of them may not be associated with SUDs in other populations. Additionally, in the 1kg-AMR-like sample, we only have GWAS of AUD and OUD; therefore, our results may be limited to these two SUDs only and not generalizable to other SUDs. Third, due to small GWAS sample sizes in the 1kg-AFR-like and 1kg-AMR-like samples, we cannot accurately estimate SNP-heritability and genetic correlations in these populations. Fourth, real-world data-based analyses were subject to unmeasured confounding and inconsistency between health insurance records and true health statuses.
Although we identified many more variants and observed greater PGS predictive power in the 1kg-EUR-like sample compared to the 1kg-AFR-like and 1kg-AMR-like samples, these differences are primarily due to the smaller sample sizes in the latter groups as well as the lack of appropriate statistical methods for analyzing admixed populations. Importantly, these findings should not be interpreted as evidence for different underlying genetic mechanisms of SUDs across these populations. Rather, our results highlight the critical and urgent need to increase the sample sizes of under-represented populations to reduce health disparities. We also emphasize that genetic factors or PGS do not determine SUD status, and environmental factors also substantially reduce or amplify the effect of genetic factors. Our findings should never be used to discriminate or to stigmatize people, or to deny access to prevention and treatment programs, especially for those from vulnerable populations. Our findings should only be used for research purposes to promote health.
In conclusion, by applying meta-analyses and retaining concordant variants across multiple SUDs, we identified a set of genes shared among different SUDs, providing new insight into their common genetic mechanisms. PGS calculated from these concordant variants had improved predictive power, particularly in the 1kg-EUR-like and the 1kg-AMR-like populations. Finally, we identified seven drugs that target SUD-share genes, highlighting their potential for repurposing in the treatment of SUDs. Together, these findings advance our understanding of the genetic architecture underlying SUDs and point to promising future prevention and treatment strategies.
Supplementary information
Acknowledgements
The All of Us Research Program is supported by the National Institutes of Health, Office of the Director: Regional Medical Centers: 1 OT2 OD026549; 1 OT2 OD026554; 1 OT2 OD026557; 1 OT2 OD026556; 1 OT2 OD026550; 1 OT2 OD 026552; 1 OT2 OD026553; 1 OT2 OD026548; 1 OT2 OD026551; 1 OT2 OD026555; IAA #: AOD 16037; Federally Qualified Health Centers: HHSN 263201600085U; Data and Research Center: 5 U2C OD023196; Biobank: 1 U24 OD023121; The Participant Center: U24 OD023176; Participant Technology Systems Center: 1 U24 OD023163; Communications and Engagement: 3 OT2 OD023205; 3 OT2 OD023206; and Community Partners: 1 OT2 OD025277; 3 OT2 OD025315; 1 OT2 OD025337; 1 OT2 OD025276. In addition, the All of Us Research Program would not be possible without the partnership of its participants. We want to acknowledge the participants and investigators of the FinnGen study. The FinnGen study is a large-scale genomics initiative that has analyzed over 500,000 Finnish biobank samples and correlated genetic variation with health data to understand disease mechanisms and predispositions. The project is a collaboration between research organizations and biobanks within Finland and international industry partners. This study was made possible, in part, with support from the Indiana Clinical and Translational Sciences Institute funded, in part by Award Number UL1TR002529 from the National Institutes of Health, National Center for Advancing Translational Sciences, Clinical and Translational Sciences Award, and the National Center for Research Resources, Construction grant number RR020128 and the Lilly Endowment. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The authors acknowledge the Indiana University Pervasive Technology Institute for providing [HPC (Big Red II, Karst, Carbonate), visualization, database, storage, or consulting] resources that have contributed to the research results reported within this paper. H.J.E. is supported by National Institutes of Health award number DA054869. D.L. and Y.L. are supported by National Institutes of Health award number AA031176.
Data availability
1kg-EUR-like Problematic alcohol use, 1kg-AFR-like alcohol use disorder GWAS (Zhou et al. [23]), and 1kg-EUR-like and 1kg-AFR-like cannabis use disorder GWAS (Levey et al. [24]): https://medicine.yale.edu/lab/gelernter/stats/. 1kg-AMR-like alcohol use disorder GWAS (Zhou et al. [23]) and opioid use disorder GWAS (Kember et al. [25]) are available through dbGaP: https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs001672. 1kg-EUR-like and 1kg-AFR-like opioid use disorder GWAS (Polimanti et al. [26]): https://figshare.com/articles/dataset/sud2020-op/14672211. 1kg-EUR-like and 1kg-AFR-like tobacco use disorder GWAS (Toikumo et al. [27]) are available by emailing the corresponding author. FinnGen Substance abuse: https://www.finngen.fi/en/access_results. The Human Brain Cell Atlas v1.0 generated by BRAIN Initiative cell census network: https://cellxgene.cziscience.com/collections/283d65eb-dd53-496d-adb7-7570c7caa443. All of Us research program: https://www.researchallofus.org/data-tools/workbench/. Indiana Biobank: https://indianabiobank.org/. Indiana biobank substance use disorder cohort is also available from dbGaP: https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs003025.v1.p1. All GWAS summary statistics from this study will be deposited to GWAS catalog: https://www.ebi.ac.uk/gwas/. All variants and their weights used to calculate PGS will be available from PGS catalog: https://www.pgscatalog.org/.
Competing interests
The authors declare no competing interests.
Footnotes
Footnote Group
Supplementary information
The online version contains supplementary material available at 10.1038/s41380-025-03294-5.
References
Untitled section
References
- 1.Degenhardt L, Charlson F, Ferrari A, Santomauro D, Erskine H, Mantilla-Herrara A, et al. The global burden of disease attributable to alcohol and drug use in 195 countries and territories, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. The Lancet Psychiatry. 2018;5:987–1012.
- 2.Deak JD, Johnson EC. Genetics of substance use disorders: a review. Psychol Med. 2021;51:2189–200.
- 3.Koob GF, Volkow ND. Neurobiology of addiction: a neurocircuitry analysis. Lancet Psychiatry. 2016;3:760–73.
- 4.Bhalla IP, Stefanovics EA, Rosenheck RA. Clinical epidemiology of single versus multiple substance use disorders: polysubstance use disorder. Med Care. 2017;55:S24–S32.
- 5.McCabe SE, West BT, Jutkiewicz EM, Boyd CJ. Multiple DSM-5 substance use disorders: A national study of US adults. Hum Psychopharmacol. 2017;32:e2625.
- 6.Kendler KS, Jacobson KC, Prescott CA, Neale MC. Specificity of genetic and environmental risk factors for use and abuse/dependence of cannabis, cocaine, hallucinogens, sedatives, stimulants, and opiates in male twins. American Journal of Psychiatry. 2003;160:687–95.
- 7.Kendler KS, Myers J, Prescott CA. Specificity of genetic and environmental risk factors for symptoms of cannabis, cocaine, alcohol, caffeine, and nicotine dependence. Arch Gen Psychiat. 2007;64:1313–20.
- 8.Kendler KS, Chen X, Dick D, Maes H, Gillespie N, Neale MC, et al. Recent advances in the genetic epidemiology and molecular genetics of substance use disorders. Nat Neurosci. 2012;15:181–9.
- 9.Hatoum AS, Colbert SMC, Johnson EC, Huggett SB, Deak JD, Pathak GA, et al. Multivariate genome-wide association meta-analysis of over 1 million subjects identifies loci underlying multiple substance use disorders. Nature Mental Health. 2023;1:210–23.
- 10.Hatoum AS, Johnson EC, Colbert SMC, Polimanti R, Zhou H, Walters RK, et al. The addiction risk factor: A unitary genetic vulnerability characterizes substance use disorders and their associations with common correlates. Neuropsychopharmacology. 2022;47:1739–45.
- 11.Johnson EO, Hancock DB, Levy JL, Gaddis NC, Page GP, Glasheen C, et al. KAT2B polymorphism identified for drug abuse in African Americans with regulatory links to drug abuse pathways in human prefrontal cortex. Addict Biol. 2016;21:1217–32.
- 12.Xiang B, Yang BZ, Zhou H, Kranzler H, Gelernter J. GWAS and network analysis of co-occurring nicotine and alcohol dependence identifies significantly associated alleles and network. Am J Med Genet B Neuropsychiatr Genet. 2019;180:3–11.
- 13.Wetherill L, Lai D, Johnson EC, Anokhin A, Bauer L, Bucholz KK, et al. Genome-wide association study identifies loci associated with liability to alcohol and drug dependence that is associated with variability in reward-related ventral striatum activity in African- and European-Americans. Genes Brain Behav. 2019;18:e12580.
- 14.Chang SH, Sun Y, Wang F, Chang XW, Zhang YJ, Jia TY, et al. Genome-wide association meta-analyses identify novel genetic risk loci and polygenic phenotype associations for heroin, methamphetamine and alcohol dependences. Clin Transl Med. 2022;12:e659.
- 15.Zhang H, Grant BF, Hodgkinson CA, Ruan WJ, Kerridge BT, Huang B, et al. Strong and weak cross-inheritance of substance use disorders in a nationally representative sample. Mol Psychiatr. 2022;27:1742–53.
- 16.Grotzinger AD, Rhemtulla M, de Vlaming R, Ritchie SJ, Mallard TT, Hill WD, et al. Genomic structural equation modelling provides insights into the multivariate genetic architecture of complex traits. Nature Human Behaviour. 2019;3:513–25.
- 17.Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics. 2010;26:2190–1.
- 18.Bulik-Sullivan BK, Loh PR, Finucane HK, Ripke S, Yang J, Schizophrenia Working Group of the Psychiatric Genomics C, et al. LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet. 2015;47:291–5.
- 19.Schoeler T, Baldwin J, Allegrini A, Barkhuizen W, McQuillin A, Pirastu N, et al. Novel biological insights into the common heritable liability to substance involvement: a multivariate genome-wide association study. Biol Psychiat. 2023;93:524–35.
- 20.Solovieff N, Cotsapas C, Lee PH, Purcell SM, Smoller JW. Pleiotropy in complex traits: challenges and strategies. Nat Rev Genet. 2013;14:483–95.
- 21.Zhu Z, Anttila V, Smoller JW, Lee PH. Statistical power and utility of meta-analysis methods for cross-phenotype genome-wide association studies. PLoS ONE. 2018;13:e0193256.
- 22.Lee PH, Anttila V, Won H, Feng Y-CA, Rosenthal J, Zhu Z, et al. Genomic relationships, novel loci, and pleiotropic mechanisms across eight psychiatric disorders. Cell. 2019;179:1469–82.e11.
- 23.Zhou H, Kember RL, Deak JD, Xu H, Toikumo S, Yuan K, et al. Multi-ancestry study of the genetics of problematic alcohol use in over 1 million individuals. Nat Med. 2023;29:3184–92.
- 24.Levey DF, Galimberti M, Deak JD, Wendt FR, Bhattacharya A, Koller D, et al. Multi-ancestry genome-wide association study of cannabis use disorder yields insight into disease biology and public health implications. Nat Genet. 2023.
- 25.Kember RL, Vickers-Smith R, Xu H, Toikumo S, Niarchou M, Zhou H, et al. Cross-ancestry meta-analysis of opioid use disorder uncovers novel loci with predominant effects in brain regions associated with addiction. Nat Neurosci. 2022;25:1279–87.
- 26.Polimanti R, Walters RK, Johnson EC, McClintick JN, Adkins AE, Adkins DE, et al. Leveraging genome-wide data to investigate differences between opioid use vs. opioid dependence in 41,176 individuals from the Psychiatric Genomics Consortium. Mol Psychiatry. 2020;25:1673–87.
- 27.Toikumo S, Jennings MV, Pham BK, Lee H, Mallard TT, Bianchi SB, et al. Multi-ancestry meta-analysis of tobacco use disorder identifies 461 potential risk genes and reveals associations with multiple health outcomes. Nature Human Behaviour. 2024;8:1177–93.
- 28.Kurki MI, Karjalainen J, Palta P, Sipilä TP, Kristiansson K, Donner KM, et al. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature. 2023;613:508–18.
- 29.Auton A, Abecasis GR, Altshuler DM, Durbin RM, Abecasis GR, Bentley DR, et al. A global reference for human genetic variation. Nature. 2015;526:68–74.
- 30.Watanabe K, Taskesen E, van Bochoven A, Posthuma D. Functional mapping and annotation of genetic associations with FUMA. Nat Commun. 2017;8:1826.
- 31.Boughton AP, Welch RP, Flickinger M, VandeHaar P, Taliun D, Abecasis GR, et al. LocusZoom.js: interactive and embeddable visualization of genetic association study results. Bioinformatics. 2021;37:3017–8.
- 32.Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res. 2010;38:e164. –e
- 33.de Leeuw CA, Mooij JM, Heskes T, Posthuma D. MAGMA: generalized gene-set analysis of GWAS data. PLoS Comput Biol. 2015;11:e1004219.
- 34.Lonsdale J, Thomas J, Salvatore M, Phillips R, Lo E, Shad S, et al. The genotype-tissue expression (GTEx) project. Nat Genet. 2013;45:580–5.
- 35.Cunningham F, Allen JE, Allen J, Alvarez-Jarreta J, Amode MR, Armean Irina M, et al. Ensembl 2022. Nucleic Acids Res. 2021;50:D988–D95.
- 36.Wang D, Liu S, Warrell J, Won H, Shi X, Navarro FCP, et al. Comprehensive functional genomic resource and integrative model for the human brain. Science. 2018;362:eaat8464.
- 37.Szklarczyk D, Gable AL, Nastou KC, Lyon D, Kirsch R, Pyysalo S, et al. The STRING database in 2021: customizable protein-protein networks, and functional characterization of user-uploaded gene/measurement sets. Nucleic Acids Res. 2021;49:D605–d12.
- 38.Sollis E, Mosaku A, Abid A, Buniello A, Cerezo M, Gil L, et al. The NHGRI-EBI GWAS Catalog: knowledgebase and deposition resource. Nucleic Acids Res. 2023;51:D977–d85.
- 39.Siletti K, Hodge R, Mossi Albiach A, Lee KW, Ding S-L, Hu L, et al. Transcriptomic diversity of cell types across the adult human brain. Science. 2023;382:eadd7046.
- 40.Benjamini Y, Hochberg Y. Controlling the false discovery rate - a practical and powerful approach to multiple testing. J R Stat Soc B. 1995;57:289–300.
- 41.Finucane HK, Bulik-Sullivan B, Gusev A, Trynka G, Reshef Y, Loh PR, et al. Partitioning heritability by functional annotation using genome-wide association summary statistics. Nat Genet. 2015;47:1228.
- 42.Speed D, Balding DJ. SumHer better estimates the SNP heritability of complex traits from summary statistics. Nat Genet. 2019;51:277–84.
- 43.Speed D, Holmes J, Balding DJ. Evaluating and improving heritability models using summary statistics. Nat Genet. 2020;52:458–62.
- 44.Lai D, Schwantes-An TH, Abreu M, Chan G, Hesselbrock V, Kamarajan C, et al. Gene-based polygenic risk scores analysis of alcohol use disorder in African Americans. Transl Psychiatry. 2022;12:266.
- 45.Lai D, Kuo SI, Wetherill L, Aliev F, Zhang M, Marco A, et al. Associations between alcohol use disorder polygenic score and remission in participants from high-risk families and the Indiana Biobank. Alcohol Clin Exp Res (Hoboken). 2023;48:283–94.
- 46.Ge T, Chen CY, Ni Y, Feng YA, Smoller JW. Polygenic prediction via Bayesian regression and continuous shrinkage priors. Nat Commun. 2019;10:1776.
- 47.Ruan Y, Lin Y-F, Feng Y-CA, Chen C-Y, Lam M, Guo Z, et al. Improving polygenic prediction in ancestrally diverse populations. Nat Genet. 2022;54:573–80.
- 48.Chang CC, Chow CC, Tellier LC, Vattikuti S, Purcell SM, Lee JJ. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience. 2015;4:7.
- 49.Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81:559–75.
- 50.Freshour SL, Kiwala S, Cotto KC, Coffman AC, McMichael JF, Song JJ, et al. Integration of the Drug-Gene Interaction Database (DGIdb 4.0) with open crowdsource efforts. Nucleic Acids Res. 2021;49:D1144–D51.
- 51.Kranzler HR, Covault J, Feinn R, Armeli S, Tennen H, Arias AJ, et al. Topiramate treatment for heavy drinkers: moderation by a GRIK1 polymorphism. Am J Psychiatry. 2014;171:445–52.
- 52.Johnson BA, Ait-Daoud N, Wang XQ, Penberthy JK, Javors MA, Seneviratne C, et al. Topiramate for the treatment of cocaine addiction: a randomized clinical trial. JAMA Psychiatry. 2013;70:1338–46.
- 53.Kosten T, Oliveto A, Feingold A, Poling J, Sevarino K, McCance-Katz E, et al. Desipramine and contingency management for cocaine and opiate dependence in buprenorphine maintained patients. Drug Alcohol Depend. 2003;70:315–25.
- 54.Agabio R, Trogu E, Pani PP. Antidepressants for the treatment of people with co-occurring depression and alcohol dependence. Cochrane Database Syst Rev. 2018;4:Cd008581.
- 55.Prochazka AV, Weaver MJ, Keller RT, Fryer GE, Licari PA, Lofaso D. A randomized trial of nortriptyline for smoking cessation. Arch Intern Med. 1998;158:2035–9.
- 56.Aryan N, Banafshe HR, Farnia V, Shakeri J, Alikhani M, Rahimi H, et al. The therapeutic effects of methylphenidate and matrix-methylphenidate on addiction severity, craving, relapse and mental health in the methamphetamine use disorder. Subst Abuse Treat Prev Policy. 2020;15:72.
- 57.Santorelli M, Miuli A, Pettorruso M, Di Carlo F, De Berardis D, Sensi SL, et al. Oral and long-acting injectable aripiprazole in severe mental illness and substance use disorder comorbidity: an updated systematic review. Curr Neuropharmacol. 2025;23:404–11.
- 58.Neyra A, Parro-Torres C, Ros-Cucurull E, Carrera I, Echarri E, Torrens M. Management of schizophrenia and comorbid substance use disorders: expert review and guidance. Ann Gen Psychiatry. 2024;23:40.
- 59.Edenberg HJ, McClintick JN. Alcohol dehydrogenases, aldehyde dehydrogenases, and alcohol use disorders: a critical review. Alcoholism-Clinical and Experimental Research. 2018;42:2281–97.
- 60.Zhou H, Sealock JM, Sanchez-Roige S, Clarke T-K, Levey DF, Cheng Z, et al. Genome-wide meta-analysis of problematic alcohol use in 435,563 individuals yields insights into biology and relationships with other traits. Nat Neurosci. 2020;23:809–18.
- 61.Walters RK, Polimanti R, Johnson EC, McClintick JN, Adams MJ, Adkins AE, et al. Transancestral GWAS of alcohol dependence reveals common genetic underpinnings with psychiatric disorders. Nat Neurosci. 2018;21:1656–69.
- 62.Kranzler HR, Zhou H, Kember RL, Vickers Smith R, Justice AC, Damrauer S, et al. Genome-wide association study of alcohol consumption and use disorder in 274,424 individuals from multiple populations. Nat Commun. 2019;10:1499.
- 63.Johnson EC, Demontis D, Thorgeirsson TE, Walters RK, Polimanti R, Hatoum AS, et al. A large-scale genome-wide association study meta-analysis of cannabis use disorder. Lancet Psychiatry. 2020;7:1032–45.
- 64.Watanabe K, Stringer S, Frei O, Umicevic Mirkov M, de Leeuw C, Polderman TJC, et al. A global overview of pleiotropy and genetic architecture in complex traits. Nat Genet. 2019;51:1339–48.
- 65.Quach BC, Bray MJ, Gaddis NC, Liu M, Palviainen T, Minica CC, et al. Expanding the genetic architecture of nicotine dependence and its shared genetics with multiple traits. Nat Commun. 2020;11:5562.
- 66.Hao L, Kraft P, Berriz GF, Hynes ED, Koch C, Korategere VKP, et al. Development of a clinical polygenic risk score assay and reporting workflow. Nat Med. 2022;28:1006–13.
- 67.Lennon NJ, Kottyan LC, Kachulis C, Abul-Husn NS, Arias J, Belbin G, et al. Selection, optimization and validation of ten chronic disease polygenic risk scores for clinical implementation in diverse US populations. Nat Med. 2024;30:480–7.
- 68.Kranzler HR, Leong SH, Naps M, Hartwell EE, Fiellin DA, Rentsch CT. Association of topiramate prescribed for any indication with reduced alcohol consumption in electronic health record data. Addiction. 2022;117:2826–36.
Associated Data
Supplementary Materials
Data Availability Statement
1kg-EUR-like Problematic alcohol use, 1kg-AFR-like alcohol use disorder GWAS (Zhou et al. [23]), and 1kg-EUR-like and 1kg-AFR-like cannabis use disorder GWAS (Levey et al. [24]): https://medicine.yale.edu/lab/gelernter/stats/. 1kg-AMR-like alcohol use disorder GWAS (Zhou et al. [23]) and opioid use disorder GWAS (Kember et al. [25]) are available through dbGaP: https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs001672. 1kg-EUR-like and 1kg-AFR-like opioid use disorder GWAS (Polimanti et al. [26]): https://figshare.com/articles/dataset/sud2020-op/14672211. 1kg-EUR-like and 1kg-AFR-like tobacco use disorder GWAS (Toikumo et al. [27]) are available by emailing the corresponding author. FinnGen Substance abuse: https://www.finngen.fi/en/access_results. The Human Brain Cell Atlas v1.0 generated by BRAIN Initiative cell census network: https://cellxgene.cziscience.com/collections/283d65eb-dd53-496d-adb7-7570c7caa443. All of Us research program: https://www.researchallofus.org/data-tools/workbench/. Indiana Biobank: https://indianabiobank.org/. Indiana biobank substance use disorder cohort is also available from dbGaP: https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs003025.v1.p1. All GWAS summary statistics from this study will be deposited to GWAS catalog: https://www.ebi.ac.uk/gwas/. All variants and their weights used to calculate PGS will be available from PGS catalog: https://www.pgscatalog.org/.