Abstract
Obesity is a major public health crisis associated with high mortality rates. Previous genome-wide association studies (GWAS) investigating body mass index (BMI) have largely relied on imputed data from European individuals. This study leveraged whole-genome sequencing (WGS) data from 88,873 participants from the Trans-Omics for Precision Medicine (TOPMed) Program, of which 51% were of non-European population groups. We discovered 18 BMI-associated signals (P < 5 × 10−9), including two secondary signals. Notably, we identified and replicated a novel low-frequency single nucleotide polymorphism (SNP) in MTMR3 that was common in individuals of African descent. Using a diverse study population, we further identified two novel secondary signals in known BMI loci and pinpointed two likely causal variants in the POC5 and DMD loci. Our work demonstrates the benefits of combining WGS and diverse cohorts in expanding current catalog of variants and genes confer risk for obesity, bringing us one step closer to personalized medicine.
Similar content being viewed by others
Introduction
In 2015, approximately 12% of adults worldwide had obesity1, and four years later, the global obesity-related deaths amounted to five million, translating to an age-standardized mortality rate of 62.6 per 100,000 individuals in 20192. Previous genome-wide association studies (GWAS) have identified hundreds of loci associated with obesity-related traits, primarily with body mass index (BMI) – a practical and widely used proxy of overall adiposity. However, most of these genome-wide screens relied on meta-analyses of imputed data, predominantly from individuals of European ancestry3,4.
Despite making some advancements toward improving ancestral diversity in GWAS, genetic ancestry-stratified analyses and multi-ancestry analyses leveraged for discovery and fine-mapping are uncommon and largely underpowered by comparison. Furthermore, follow-up investigations for known BMI loci identified in European-ancestry populations are insufficiently conducted to evaluate the generalizability of these loci. As such, the majority of BMI-risk variants are common variants (minor allele frequency [MAF] > 5%) in primarily European-ancestry populations, most of which exhibit small effect sizes. While these index variants collectively explain less than 5% of the total phenotypic variation in BMI5, it is estimated that as much as 1/5 of the phenotypic variance can be captured by common variants across the entire genome5, leaving low-frequency and rare variants (MAF ≤ 5%) with potentially large effects to be explored6.
Whole-genome sequencing (WGS) outperforms genotyping arrays in capturing low and rare frequency variants, as demonstrated in a recent study where researchers revealed that the heritability of BMI estimated using WGS data was comparable to the pedigree-based estimates, h2 ≈ 0.407. Thus, the discrepancy between phenotypic variance explained by genetic variations in GWAS compared to the expected heritability may be due to the use of imputed genotypes rather than directly sequenced variations. Causal variants or SNPs in known loci may not be represented on 1000 Genomes panels or not well imputed from reference data because of differences in linkage disequilibrium (LD) across populations. To address this limitation, we conducted WGS association analyses to identify rare, low-frequency, and genetic ancestry-specific genetic variants associated with BMI, using data from the Trans-Omics for Precision Medicine (TOPMed) Program8, which is the most racially and ethnically diverse WGS program to date, as well as the Centers for Common Disease Genomics (CCDG) Program9.
Results
Single-variant analyses
Our study population was racially, ethnically, geographically, and ancestrally diverse. We analyzed a multi-population sample of 88,873 adults from 36 studies in the Freeze 8 TOPMed and CCDG programs (Fig. 1, Supplementary Fig. 1, Supplementary Data 1–5). Among the 90 million SNPs included in the multi-population analysis, 86% (N = 77,178,487) were rare SNPs with a study-wide MAF of 0.5% <MAF ≤ 1%, and 6% (N = 5,542,150) were low-frequency (1% <MAF ≤ 5%) SNPs. In the multi-population unconditional analysis, we identified 16 loci that reached the prespecified genome-wide significance threshold of P < 5 × 10−9 (Table 1, Fig. 2, Supplementary Figs. 2 and 3), including one low-frequency (MAF = 4%) and 15 common (MAF 14%–50%) tag SNPs. In general, the low-frequency variant in our primary discovery showed a stronger effect than the common variants, with an estimated effect 2.14 times larger than the average common variants (0.078 vs. 0.037 on average). Of these 16 loci, 15 were in known BMI-associated regions, and one novel locus was identified on chromosome 22 harboring a low-frequency index SNP near MTMR3 (Myotubularin-Related Protein 3; rs111490516; MAF = 4%, β = 0.078, SE = 0.013, P = 4.52 × 10−9; Table 1). The MAF of this MTMR3 locus varied widely across population groups, with the highest MAF observed in the African (13%) and Barbadian (13%) population groups, while it ranged from 0% to 5% in other population groups (Supplementary Data 5).
A Pairwise scatter plots of the first three principal components (PCs) by population group. B The number and proportion of participants by population group. Our study population was composed of 88,873 participants from 15 population groups, 51% of which are non-European.
A Manhattan plot of multi-population, single-variant analysis (N = 88,873 individuals). The novel locus (MTMR3) is highlighted in red. Previously reported BMI loci are in dark beige. The horizontal dashed line indicates the genome-wide significance threshold two-sided P = 5 × 10−9, to account for multiple testing. B Scatterplot showing the minor allele frequency compared to the absolute value of the estimated effect of the index variant at each significant locus. All effect estimates are from the primary analysis conducted across all population groups. Previously reported loci are highlighted in blue, while the novel locus is in red; circles represent the most significant variant at each locus, and triangles show newly reported secondary signals within known loci.
In the two population-specific analyses, 10 association signals reached genome-wide significance (Supplementary Data 6, Supplementary Figs. 4–7). All 10 association signals were identified in the multi-population analysis, including two of these loci with the same index SNPs in at least one population, SEC16B in Africans and FTO in Europeans. For these same two loci, each population-specific analysis also revealed a distinct lead variant compared to the multi-population analysis; however, they were in high LD with (R2 = 0.95 and R2 = 1.00, according to TOP-LD10; Supplementary Data 6) and within 30 kb of the multi-population lead SNPs. Notably, the novel locus in MTMR3 achieved significance exclusively in the African group. While the most significant SNP in the African population group (rs73396827) differed from that in the multi-population analysis (rs111490516), the two were in strong LD in the TOPMed African population (R2 = 1.00). Both of these SNPs were fixed in the European group (MAF = 0%). Our sensitivity analysis illustrates the robustness of our findings as we identified the same three significant association signals (P < 5 × 10−9) and lead SNPs as in the main analysis (Supplementary Data 7). Additionally, there was no substantial change in the effect estimate (directionally consistent and sensitivity |Beta| > 90% original |Beta|). In the European group analysis, one SNP in the ALKAL2 locus on chromosome 2 (rs62107261, β = −0.102, SE = 0.016, P = 2.08 × 10−10, MAF = 5%) was not in LD with the corresponding lead variant in the multi-population analysis (R2 = 0.00, as calculated in the analysis subset), but was a known independent secondary signal at this locus11. The remaining SNPs were in the proximity to and in LD with the index SNPs in the corresponding loci from the multi-population analysis.
Replication
The replication sample sizes ranged from 3213 in MyCode to 80,730 in MVP (Supplementary Data 8). In the six replication studies of Black, African, and African American participants, rs111490516 and rs73396827 in MTMR3 displayed high LD (R2 = 1.00) (Supplementary Data 8). Their MAFs are nearly identical across studies and ranged from 11% to 13%, aligning with the 13% observed in our African and Barbadian groups (Supplementary Data 8). We replicated both variants, demonstrating directionally consistent associations with BMI across the replication studies. In MVP and REGARDS, particularly, both variants reached statistical significance at P < 0.05. We observed a 68% reduction in the estimated effects when meta-analyzing across replication studies (rs111490516: β = 0.025, SE = 0.007, P = 1.92 × 10−4, MAF = 12%; rs73396827: β = 0.026, SE = 0.007, P = 1.50 × 10−4, MAF = 12%), compared to the discovery analysis (Fig. 3, Supplementary Data 8, Supplementary Fig. 8). In the meta-analysis of up to 202,675 individuals from both discovery and replication studies, rs111490516 and rs73396827 both had an estimated effect of 0.36 with a SE of 0.006 and P-values of 2.40 × 10−9 and 1.60 × 10−9, respectively (Fig. 3, Supplementary Data 8).
The forest plot, centered on effect estimates with 95% confidence intervals, is oriented on the BMI-increasing allele. Effect estimates are provided as standard deviation in BMI per allele. Standard errors and P-values of the effect estimates are provided in Supplementary Data 8.
Functional annotation
To gain a better understanding of the potential functional consequence of the MTMR3 locus, we used publicly available information from Ensembl Variant Effect Predictor (VEP)12, FORGEdb13, and ENCODE14,15 to annotate nearby variants. Of the 54 variants in high LD with our lead SNP, most were intronic or nearby MTMR3 (Supplementary Data 9). Of these, four variants had a moderate CADD (combined annotation dependent depletion) score (scaled CADD > 10) with rs73394881 having the highest relative CADD score16, three of which lay within a possible enhancer (rs73396896, R2 = 0.884; rs73394881, R2 = 0.889; rs74832232, R2 = 0.889). There is an abundance of evidence for regulatory action in this region for variants in high LD with our lead SNP. Although our top associated SNP does not directly overlap potential candidate cis-regulatory elements (cCREs) based on ENCODE, there is limited data of overlapping cCREs from RoadMap data available on FORGEdb (Supplementary Fig. 9, Supplementary Data 9). Three genes were implicated across these databases, including MTMR3, MTFP1 (Mitochondrial Fission Process Protein 1), and ASCC2 (Activating Signal Cointegrator 1 Complex Subunit 2). Five SNPs were implicated in potential regulation for MTMR3 and/or ASCC2 across multiple tissues based on Activity-By-Contact (ABC) data. Additionally, 35 SNPs were significant cis-eQTLs with MTFP1 in Nerve Tibial cells. Three SNPs exhibited Zoonomia PhyloP scores indicative of accelerated evolution (PhyloP ≤ −2.270), including our tag SNP, rs111490516, and two nearby SNPs, rs57349783 and rs2107673; meanwhile, three exhibited scores indicative of significant cross-species conservation (PhyloP ≥ 2.270), including rs73396896, rs6006286, and rs7339488117.
Conditional analyses
Conditional analysis using the most associated variant at each locus revealed two significant secondary signals after multiple testing correction (Table 1, Supplementary Data 10, Supplementary Fig. 10). These included a known BMI-associated index variant on chromosome 2 (rs62107261, β = −0.097, SE = 0.014, P = 2.06 × 10−12, near ALKAL2)11, which was also the most significant variant at this locus in the European group analysis (Supplementary Data 6). We further identified rs78769612 on chromosome 18 (β = −0.100, SE = 0.019, P = 2.17 × 10−7, near MC4R). Although both secondary SNPs were in known BMI-associated loci, rs78769612 near MC4R was a new index variant.
We additionally assessed independence for the top variants in known loci by conditioning on all previously reported index variants by chromosome5,11,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33, and highlighted previously reported index variants located within 500 kb of our top locus SNP (Supplementary Data 11). Two SNPs, rs2206277 in TFAP2B and rs3838785 in BDNF, remained significant after multiple test corrections, indicating potentially novel signals in known loci (Supplementary Data 11). The novel index variant from the internal conditional analysis, rs78769612 near MC4R, was not robust to this treatment, suggesting that this novel variant was not independent of known BMI variants. The LD matrix plots highlighted low LD (R2 range 0.018–0.342) between our top SNP at the BDNF locus, rs3838785, and previously published lead variants within 500 kb (Supplementary Fig. 11). Although our top SNP, rs2206277, in the TFAP2B locus was conditionally independent of previously published BMI-risk SNPs (β ≥ 90% of the unconditioned β and P < 6.25 × 10−3), this SNP was in high LD with two nearby published SNPs (R2 = 0.822 for rs987237 and R2 = 0.793 for rs72892910).
Aggregate-based testing
We did not identify any novel gene regions through association tests at the genome-wide level (P < 5 × 10−7) when aggregating variants with MAF ≤ 1%. Nevertheless, we successfully replicated previous gene-based associations with the well-known MC4R gene (P = 8.47 × 10−8), with 110 alleles across 37 sites within coding regions, enhancers, and promoters for MC4R. The MC4R locus was also identified in single-variant analyses. We also examined the 16 genes reported by Akbari et al.34 with an exome-wide significant association with BMI (Supplementary Data 12). Although only MC4R reached genome-wide significance in our aggregate-based testing, three additional genes – ROBO1, GPR151, and ANO4 – showed nominal significance (P < 0.05) in our SMMAT test.
Fine-mapping
To pinpoint the most probable causal variant(s) underlying each of the 16 loci, we subsequently performed fine-mapping using PAINTOR35. Assuming one causal variant per locus, the index variants were the most likely causal variants in 14 loci, with posterior probability (PP) ranging from 0.02 and 1.00, and seven of them had a PP above 0.50 (Supplementary Data 13, Supplementary Fig. 12). Two intronic index variants, rs2307111 in POC5 and rs1379871 in DMD, were particularly noteworthy with PP exceeding 0.98. In contrast, variants with the highest PP in ADCY3 and ZC3H4 were not the reported index variants, although the highest PP for the ADCY3 locus was below 0.50 and thus not likely the causal variant underlying this signal. In the ZC3H4 locus, the highest PP variant (rs55731973, PP = 0.77) was intronic, located in the 5′ UTR or upstream of ZC3H4 depending on alternative transcripts, and resided in probable enhancer regions. Additionally, this variant was a significant cis-eQTL for SAE136, another nearby downstream gene. Two of our loci identified potential secondary signals following conditional analysis, the TMEM18/ALKAL2 and MC4R loci, but both secondary signals were >100 kb from our index variant.
Of the seven variants with PP above 0.50, all have evidence of moderate to high impact on a nearby gene. All had FORGEdb13 scores of 5 or above, and six had CADD scores greater than 1516 for at least one transcript (Supplementary Data 9). For example, the causal variant implicated in our fine-mapping for the FTO locus, rs1421085, is intronic or within an enhancer for FTO, has a FORGEdb score of 9, and scaled CADD score of 19.58. Multiple studies on this known obesity-risk variant have confirmed this as a causal variant in the pathogenesis of obesity37,38,39,40. Also, rs55731973, in the ZC3H4 locus, while not our index SNP, has a FORGEdb score of 10 (the highest score) and a scaled CADD score of 18.84 due to multiple lines of evidence that it sits in an active regulatory site across multiple tissues. None of the seven exhibited a CATO (Contextual Analysis of TF Occupancy) score >0.1, indicative of strong cross-tissue evidence of transcription regulation41, a conservative metric for identifying high-confidence regulatory variants. Four of the seven SNPs exhibited high Zoonomia PhyloP scores indicative of significant cross-species conservation (PhyloP ≥ 2.270), including rs1379871, rs2307111, rs55731973, and rs142108517.
PheWAS
To explore potential novel pleiotropy, we conducted association tests between the tag variant from our novel locus, rs111490516, and 538 PheCodes available in the MyCode and BioMe studies. No PheCode was significantly associated with rs111490516 following multiple test corrections (P < 9.3 × 10−5). However, PheCode 327.3 (Sleep Apnea) and 327.32 (Obstructive Sleep Apnea) ranked among the top associated PheCodes (P < 0.001) (Supplementary Data 14, Supplementary Fig. 13). Perhaps not coincidentally, obesity is one of the strongest risk factors for sleep apnea42.
Discussion
By leveraging WGS data from a large multi-population study, we identified and replicated one novel low-frequency BMI variant in MTMR3 specific to the diversity of our sample. We also identified two common secondary signals in known BMI loci, supported gene-based associations for MC4R, and refined resolution in multiple loci by prioritizing candidate SNPs with high PP. Our discovery of the novel BMI-associated variant emphasizes the importance of studying diverse populations, which could further refine and expand the catalog of genes and variants that confer risk for obesity and potentially other disease traits.
The novel MTMR3 variant, rs111490516, was most common in our African and Barbadian population groups (MAF = 13%) and of moderate frequency in our Dominican population group (MAF = 5%). We further replicated this association in study samples of similar population backgrounds. Yet, previous GWAS of BMI focusing on African ancestry individuals failed to identify a significant association in this region. It is not available for lookup in the most recently published MVP BMI GWAS43, although included in our replication with the latest MVP data release. In one of the largest GWAS meta-analyses of imputed genotype data in African ancestry individuals with summary data available publicly, which was conducted by the African Ancestry Anthropometry Genetics Consortium (AAAGC, N up to 42,751)23, this variant was directionally consistent and suggestively associated (β = 0.042, P = 1.80 × 10−4, MAF = 12%)23. Similarly, in our replication analysis of 113,802 individuals with imputed genotypes, rs111490516 was suggestively significant (β = 0.025, P = 1.92 × 10−4, MAF = 12%). In the African-specific analysis, although rs11490516 was genome-wide significant (β = 0.087, SE = 0.015, P = 1.82 × 10−9), rs73396827 was the top hit at this locus. These two SNPs further exhibited absolute LD, as well as consistent directions of effect, SEs, and P-values across replication studies and in the meta-analyses. Therefore, the lack of discovery in prior publications may not be due to insufficient power. As indicated by our fine-mapping results and the potential regulatory role of nearby variants in this novel locus, our index SNP is likely not causal but could be in LD with a causal SNP and also poorly captured in studies relying on imputation. In other words, the causal variant underlying this locus may be nearby, less frequent, and on an LD block more frequent in a population poorly represented in other imputation reference panels but well represented in our WGS and highly diverse sample (e.g., Caribbean admixed individuals). The non-European-ancestry populations particularly fall short in imputation performance due to their persistent underrepresentation in reference panels and overestimate of imputation accuracy44. In this case, one would require sequencing data in a large sample size with the relevant haplotype to detect a significant association that was not able to be identified with imputation in a similar number of people. Unfortunately, we are unable to test this hypothesis in our data due to data access constraints. Thus, future studies with WGS in study populations with genetic similarity and functional follow-work are needed to further narrow in on the causal variant(s) underlying this association signal.
The SNP rs111490516 lies in an intron of the MTMR3 (myotubularin-related protein 3) gene, with limited evidence of involvement in regulatory or functional protein activity. Other variants mapped to MTMR3 have been associated with obesity-related traits in GWAS. In a study of 155,961 healthy and medication-free UKBB participants, rs5752989 near MTMR3 was associated with fat-free mass (β = 0.115, P = 8.00 × 10−9, allele G frequency = 43%)45. In a meta-analysis of up to 628,000 BioBank Japan (BBJ), UKBB, and FinnGen (FG) participants, the same SNP was associated with body weight (β = -0.010, P = 3.86 × 10−8, allele A frequency ranged from 51% in FG to 86% in BBJ)46. However, this SNP (rs5752989, chr22:29969791) is not in LD of our top associated SNP in this region in our study population (R2 = 0.03), and thus likely an independent signal.
While there is limited knowledge on the biological implications of the lead variant, there is an abundance of evidence for a regulatory role for SNPs in high LD with our lead SNP and multiple lines of evidence supporting a role in obesity at this locus. We utilized the Ensembl VEP, FORGEdb, and ENCODE databases to explore the predicted functional consequences of our novel locus. Our lead variant is intronic to MTMR3, and there is evidence linking SNPs in high LD with the regulation of MTMR3, MTFP1, and ASCC2. The primary cellular function of MTMR3 relates to the regulation of autophagy47. Although there is no direct evidence linking MTMR3 to obesity, previous studies have established a connection between MTMR3 and related cardiometabolic traits. MTMR3 was associated with LDL cholesterol (P = 1 × 10−8) in a GWAS meta-analysis of European, East Asian, South Asian, and African ancestry individuals48. A potential mechanism was proposed later suggesting MTMR3 may mediate the association between miRNA-4513 and total cholesterol49. Furthermore, pyruvate dehydrogenase complex-specific knockout mice with high-fat diet-induced obesity also exhibited increased blood glucose and higher expression levels of MTMR350. ASCC2 has no known role related specifically to obesity. However, like previously implicated BMI-associated genes5, it is a ubiquitin-binding protein involved in transcriptional regulation and DNA repair51,52,53,54. There is strong evidence for a role of MTFP1 in the regulation of adipogenesis from animal models. For example, liver-specific knockout of MTFP1 in mice provides evidence for protection against weight gain and ensuing diseases of metabolic dysregulation (e.g., diabetes and non-alcoholic fatty liver disease)55. The protection against weight gain appears specific to lipid-rich diets through and less pronounced in exposure to carbohydrates. Furthermore, the knockdown of MTFP1 in adipocytes from sheep increased adipogenesis. Decreased expression of MTFP1 was linked to increased expression of PPARG (Peroxisome Proliferator-Activated Receptor Gamma) and LPL (Lipoprotein Lipase), two known obesity-related genes important for adipogenesis, lipid metabolism, and ultimately energy homeostasis. Also, differential expression of MTFP1 was observed during increased adipogenesis and fat deposition in the tail of developing sheep56.
The use of WGS, coupled with the inclusion of non-European populations, improved fine-mapping resolution, as has been shown previously32. While there have been multiple attempts to fine-map previously identified BMI loci5,20,33, no previous peer-reviewed and published study has successfully identified BMI-risk variants of high confidence at the POC5 and DMD loci. By applying a Bayesian fine-mapping approach, we reduced associated signals to 95% credible sets of two likely causal SNPs. Functional annotation revealed that one of them, rs2307111, was a benign missense variant in POC5 (NP_001092741.1:p.His36Arg) according to ClinVar57,58, while the other is an intron variant in the promotor region of DMD. Nevertheless, these two variants were also considered high-confidence causal variants (PPrs2307111 = 0.96 in UKBB, PPrs1379871 = 0.99 in both UKBB and FG) in a recent preprint of a joint analysis of three biobanks (UKBB, FG, BBJ)59. Notably, unlike in Kanai et al. where the PP appeared to be driven by the Europeans (for rs2307111: PPUKBB = 0.96, PPBBJ = 0.12, PPFG = 0.01; for rs1379871: PPUKBB = 1.00, PPFG = 1.00), the effect alleles in our study were observed in high proportions across many non-European population groups (Supplementary Data 5). Beyond these two high confident SNPs, five additional SNPs in as many loci were identified in our fine-mapping analysis with a PP > 0.5, all with moderate to high evidence of a regulatory or functional role related to a nearby gene. Given that one of these SNPs, rs1421085 in FTO, has been successfully confirmed as a causal variant at this locus37,38,39,40, additional variants highlighted in our fine-mapping analysis warrant consideration in future functional studies. While the use of population-matched LD is ideal for trans-population fine-mapping, such data remain scarce in non-European populations60. Although trans-population fine-mapping strategies assuming multiple causal variants are considered superior61, PAINTOR may yield unreliable results under certain conditions, even when assuming a single variant62. Even though we are encouraged that Kanai et al.59, using FINEMAP63 and SuSiE64, identified the same two variants as high-confidence causal variants, cross-validation of results using other fine-mapping approaches remains necessary.
In addition to our novel findings, 17 of the 18 identified variants reside in previously reported BMI-associated loci, highlighting the generalizability of the genes underlying BMI across populations, including SEC16B, TMEM18, ETV5, GNPDA2, BDFN, and MC4R5,21,33,65. Three of the loci harbor genes implicated in severe and early-onset obesity – ADCY3, BDNF, and MC4R4. We also consistently identified multiple association signals of high effect in MC4R, which is a well-established monogenic obesity gene, through our discovery analysis, internal conditional analysis, and rare variant aggregate analysis. Despite not identifying novel SNPs in MC4R that are independent of known BMI-associated SNPs, we replicated a secondary signal in this gene, rs2229616, a rare missense variant previously reported in individuals of European ancestry by Speliotes et al.66. In addition to MC4R, three other genes – ROBO1, GPR151, and ANO4 – of the exome-wide significant genes identified in Akbari et al.34 showed nominal significance in our rare variant aggregate analysis. This lends further support to the generalizability of these genes across populations, given that 85% were of European ancestry and 15% of admixed American ancestry in Akbari et al., compared to our more diverse cohort with 49% European and 51% other populations. We did not replicate 12 of the gene-based findings from Akbari et al. The previous study had a nearly seven-fold larger sample size compared to the current study, included related individuals increasing the opportunity for multiple copies of rare variants, and implemented alternative methods for gene-based analysis and variant binning. Therefore, it is likely a combination of power due to differences in sample size, underlying variant selection, difference in methods for gene-based analysis, and potential winner’s curse that contributed to the lack of validation of their findings in the current study.
While our study included a large sample size of diverse populations and leveraged high-quality WGS data from well-characterized and harmonized cohorts, our results should also be interpreted with the following limitations. First, although our study is large compared to other harmonized and sequenced data samples, the total study size is relatively modest compared to existing GWAS meta-analyses of common variants using imputed genotype data. Moreover, rare variants, such as those analyzed in our study, may require even larger sample sizes for novel discoveries. Even though our study is among the most racially, ethnically, and ancestrally diverse yet conducted, the European population group still represented 49% of our participants. On the other hand, diversity can contribute to added heterogeneity of effect sizes for common and rare variants, potentially limiting discovery in the multi-population analysis. We sought to overcome this limitation by allowing for heterogeneous residual variances across population groups and examining population-stratified results when sample sizes were adequate. Notably, all our genome-wide significant loci from population-stratified analyses were also captured in the multi-population analysis, likely owing to our considerations of heterogenous effects, self-identity (population groups), and genetic ancestry (genotype principal components [PCs]). As has been shown by others32,67,68, this underscores the importance of conducting multi-population analysis using appropriate methods that account for heterogeneity and minimize the risk of inflation or missed detection of loci that may vary in MAF or phenotypic effects across populations.
In summary, our study demonstrates the power of leveraging WGS data from diverse populations for new discoveries associated with BMI. As we enter the era of incorporating GWAS-based risk models in clinical practice, it is critical that we continue to diversify the data collected and analyzed in genomic research. Failure to do so risks further exacerbating health disparities for public health crises such as obesity. Ultimately, our study brings us one step closer to understanding the complex genetic underpinnings of obesity, translating these leads into mechanistic insights and developing targeted preventions and interventions to address this global public health challenge.
Methods
Study population and phenotype
Our study population was racially, ethnically, geographically, and ancestrally diverse. We analyzed a multi-population sample of 88,873 adults from 36 studies in the freeze 8 TOPMed and CCDG programs (Fig. 1, Supplementary Data 1). TOPMed Program individual-level data is available through Google and AWS cloud services following NIH dbGap approval. Details on gaining access are found on the TOPMed website (see Data Availability Statement). They belonged to 15 population groups, reflecting the way participants self-identified in each study. For individuals who had unreported or non-specific population memberships (e.g., “Multiple” or “Other”), we applied the Harmonized Ancestry and Race/Ethnicity (HARE) method69 to infer their group memberships using genetic data, excluding these individuals from the training step. This imputation was applied to 8015 participants (9% of the overall population), assigning each to one of the existing population groups based on the group with the highest probability of membership. All other participants remained in the population group assigned based on their self-reported race/ethnicity/population group. In this way, our study population groups were defined based on a combination of self-reported identity and the first nine genetic PCs (Fig. 1, Supplementary Fig. 1, and Supplementary Data 1). The decision to use nine PCs was informed by the elbow method and scree plots.
The 15 population groups were labeled by their self-identified or primary inferred population group (e.g., predominantly African ancestry/admixed African/Black participants were labeled as “African”). Sample sizes for these groups ranged from 341 to over 43,000 as follows: African (N = 22,488), Amish (N = 1106), Asian (N = 1241), Barbadian (N = 248), Central American (N = 776), Costa Rican (N = 341), Cuban (N = 2128), Dominican (N = 2046), European (N = 43,434), Han Chinese (N = 1787), Mexican (N = 4265), Puerto Rican (N = 4991), Samoan (N = 1274), South American (N = 695), and Taiwanese (N = 2053). We refer to analyses involving all 15 population groups as multi-population analyses and group-specific analyses by their primary population group.
Among the 88,873 participants, 53,109 (60%) were female, and 45,439 (51%) were non-European. The mean (SD) age of the participants was 53.5 (15.1) years. Additional descriptive tables of the participants are presented in Supplementary Data 2–4. BMI was calculated by dividing weight in kilograms by the square of height in meters. Participants were excluded from analyses if less than 18 years of age, had known pregnancy at the time of BMI measurement, had implausible BMI values (above 100 kg/m2 without corroborating evidence), or did not provide appropriate consent. The mean (SD) of BMI varied by study, ranging from 23.4 (3.1) in GenSALT to 33.9 (7.8) in DHS (Supplementary Data 2), and by population group, ranging from 23.4 (3.1) in Han Chinese to 33.7 (6.8) in Samoans (Supplementary Data 3).
TOPMed WGS
In TOPMed, ~30× WGS was conducted using Illumina HiSeq X Ten instruments at six sequencing centers8. At the Center for Statistical Genetics at the University of Michigan, TOPMed sequence data were mapped to the GRCh38 human genome reference sequence in a manner consistent with the joint CCDG/TOPMed functionally equivalent read mapping pipeline70. Joint genotype calling on all samples in Freeze 8 used the GotCloud pipeline71. Variants were filtered using a Support Vector Machine (SVM) implemented in the libsvm software package. Sample-level quality assurance steps included concordance between annotated and genetic sex, between prior SNP array genotyping and WGS-derived genotypes, and between observed and expected relatedness and pedigree information from cleaned sequence data. These details regarding the laboratory methods, data processing, and quality control are also described on the TOPMed website (https://topmed.nhlbi.nih.gov/topmed-whole-genome-sequencing-methods-freeze-8).
Common variant association analysis
We performed a multi-population WGS association analysis of BMI using GENESIS72 on the Analysis Commons (http://analysiscommons.com)73 computation platform. GENESIS was chosen due to its analytical flexibility in relationship to allowing for heterogeneity of effect by population group72, an option well-suited to the demographic and genetic background of our study population. Analyses were performed using linear mixed models (LMM). To improve power and control for false positives with a non-normal phenotype distribution, we implemented a fully adjusted two-stage procedure for rank-normalization when fitting the null model74. The first model was fit by adjusting BMI for age, age squared, sex, study, population group, and genetic ancestry-representative PCs generated using PC-AiR75, sequencing center, sequencing phase, and project. A 4th-degree sparse empirical kinship matrix (KM) computed with PC-Relate67 was included to account for genetic relatedness among participants. We also allowed for heterogeneous residual variances across sex by population group (e.g., female European), as it has previously been shown to improve control of genomic inflation76. Residuals from the first model were rank-normal transformed within population group and sex strata. The resulting transformed residuals were used to fit the second-stage null model, allowing for heterogeneous variances by the population group and sex strata and accounting for relatedness using the kinship matrix. Variants with a MAF of at least 0.5% were then tested individually. Due to the large number of variants tested (N = 90,142,062) in the multi-population analysis, we adopted a significance threshold of 5 × 10−9 as has been used previously77. This approach maximizes participant inclusion, thereby maximizing the sample size in our discovery cohort to increase statistical power and avoid the misinterpretation of group-specific effects in underpowered strata. By including ancestrally diverse populations, we leverage differences in LD across populations, which has been shown to help identify novel loci, narrow down causal variants, and improve the variance explained in models32,68,76. For quantitative traits like BMI, multi-population analyses that account for heteroscedasticity of genetic effects across population groups have proven effective in increasing study power and reducing genomic inflation32,76. Lastly, and most importantly, a pooled approach aids in decreasing health disparities by identifying loci that generalize across populations.
Additionally, group-specific analyses were conducted in the two largest population groups, European and African, to determine whether a particular group with a large sample size is driving the observed association signals. To address any concerns over potential residual confounding due to population substructure in our African group-specific GWAS, especially for our novel association that differed in allele frequency across population groups, we conducted a sensitivity analysis using group-specific PCs. The sensitivity analysis used 10 African group-specific PCs calculated in the same individuals as used in the pooled analysis and the same covariates as before.
Replication analysis
For the novel single-variant association identified in the MTMR3 locus from our discovery analyses that is largely driven by the African population group, we requested replication specifically in Black, African, and African American participants from six independent cohorts (Ntotal = up to 113,802): BioMe BioBank Program78, Million Veteran Program (MVP)43,79, Multiethnic Cohort (MEC)80, MyCode Community Health Initiative Study (MyCode)81, REasons for Geographic And Racial Differences in Stroke (REGARDS) study82, and UK Biobank (UKBB)83,84. Phenotypes were developed and analyses were carried out under the same protocol as outlined above. We subsequently conducted inverse variance weighted fixed effects meta-analysis in STATAv15.185, using study-specific summary results. Additional details on the parent study design for each replication study are included in the Supplementary Note.
Functional annotation
To gain a better understanding of the potential functional consequence of the MTMR3 locus, we used Ensembl VEP12 and FORGEdb13 to annotate all variants in high LD with our top SNP (R2 > 0.8 in the African population group using TOP-LD10). Additionally, we looked for variants that overlap with potential cCREs within 100 kb of the lead index variants using visual inspection of regional association plots, including LD information. The resources included signatures of promoters, enhancers, and chromatin accessibility (i.e., markers of histone modification, DNase hypersensitivity, CTCF binding, etc.) from BMI-relevant tissues, including blood, brain, and liver from ENCODE14,15.
Conditional analysis
To identify loci harboring multiple independent signals, we performed stepwise conditional analyses on the most significant signal within 500 kb of our index variant. The significance threshold for secondary signals was determined by Bonferroni correction for the number of variants across all regions tested, P = 5.96 × 10−7 (P < 0.05/83,928 SNPs with MAF > 0.5% within 500 kb of the 16 index SNPs). Variants passing the significance threshold after the first round were further conditioned on the top variant in the locus after the first round of conditioning to identify potential third signals within each locus using the same threshold.
To determine whether association signals in known loci were independent of known signals, we performed conditional analyses using previously published index variants5,11,18,19,20,21,22,23,24,26,27,28,29,30,31,32,33,86. Specifically, for each lead SNP that is not a previously reported BMI-associated index variant, we conditioned on all known SNPs on each chromosome. Given that these are potential new signals in regions known to influence BMI, index variants were considered independent if the estimated effect (β) value remained ≥90% of the unconditioned β value and P < 6.25 × 10−3 (0.05/8 loci tested). For visual inspection and calculation of LD in reference panels used for imputation in previous GWAS studies, LDlink was used to calculate pairwise LD between potentially independent signals in known loci and produce LD heatmaps using the 1000 Genomes Global reference panel87.
Rare variant aggregate association analysis
Rare variants with a MAF ≤ 1% were tested in aggregate by gene unit across studies in the multi-population analysis. Variants were grouped into gene units in reference to GENCODE v28, including both coding variants and variants falling within gene-associated noncoding elements. Coding variants included high-confidence loss of function variants annotated by LOFTEE88 (Ensembl VEP LoF = HC), missense variants (MetaSVM score >0), and in-frame insertions/deletions or synonymous variants (FATHNMM-XF coding score >0.5). In addition to coding variants, we included variants falling within the promoter of each transcript tested. Promoter regions were defined as falling in the 5 kb region 5′ of the transcript and also overlaying a FANTOM5 Cage Peak89. In order to identify regulatory elements likely to be acting through the tested gene, we leveraged the GeneHancer database90. GeneHancer identifies enhancer regions and associates them with the specific genes they are likely to regulate, allowing us to aggregate regulatory regions by the likely target gene. GeneHancer regions were limited to the top 50% scored regions, and variants falling in these regulatory elements were further filtered to those most likely to have a functional impact (FATHMM-MKL noncoding score >0.75). Variants aggregated to gene units were tested using variant-set mixed model association tests (SMMAT)91. Variants were weighted inversely to their MAF using a beta distribution density function with parameters 1 and 25. Genes were considered significantly associated after Bonferroni correction for the number of genes analyzed (P < 5 × 10−7). These annotations were selected by the TOPMed Data Coordinating Center (DCC) as part of the centralized and harmonized annotations and were chosen to focus on annotation that was previously shown to have high agreement with other annotation resources and high prediction accuracy92.
Fine-mapping
In order to identify candidate functional variants underlying association regions, we performed fine-mapping analyses in our multi-population GWAS single-variant association summary statistics, using the program PAINTOR35, which integrates the association strength and genomic functional annotation. We used the annotation file from aggregate-based testing described above under “Rare Variant Aggregate Association Analysis” to identify deleterious coding variants, variants within GeneHancer regions, and variants within gene promoter regions. We restricted this analysis to variants located within ±100 kb of the locus index variants. We calculated LD using our analysis subset of the TOPMed data. As PAINTOR may be sensitive to the misspecification of the number of causal variants62,93,94, limiting our ability to interpret findings in the absence of evidence for more than one signal, we assumed one causal variant per locus, unless evidence of independent secondary signals within the 100 kb window was identified following conditional analysis, in which case we allowed for additional causal variants per locus. Restriction to a 100 kb ± window was applied due to the computational burden of WGS data and because, for most loci, LD decays at >25 kb95,96. While we extend our fine-mapping locus to 100 kb to allow for potential LD beyond 25 kb, one limitation of this approach is that we may still miss potentially causal variants that are >100 kb from our index SNP, including secondary signals. To gain a better understanding of the potential functional consequence of likely causal variants within each locus, we used Ensembl VEP12 and FORGEdb13 to annotate all variants with PP > 0.5.
PheWAS
To identify potential novel phenotypic associations with newly discovered variants, we performed a phenome-wide association (PheWAS) in the MyCode, a hospital-based population study in central and northeastern Pennsylvania81, and in the Charles Bronfman Institute for Personalized Medicine’s BioMe BioBank Program located in New York City97; both studies had genetic data linked to electronic health records (EHR). ICD-10-CM and ICD-9-CM codes were mapped to unique PheCodes using the Phecode Map v1.298 from the EHR. Cases were defined if individuals had two or more PheCodes on separate dates, while controls had zero instances of the relevant PheCode. We performed association analyses on PheCodes with N ≥ 20 cases and 20 controls using logistic regressions, adjusting for current age, sex (for non-sex-specific PheCodes), and the first 15 PCs in BioMe and 10 PCs in MyCode calculated from genome-wide data, and assuming an additive genetic model using the PheWAS package99 in R.Given the potential population-specific association of our novel locus, PheWAS were restricted to African Americans in each study: in MyCode, African Americans were identified through electronic health records (a combination of self-report and clinician-reported race/ethnicity), while in BioME the identification was based exclusively on self-report. We restricted our analyses to unrelated individuals up to the 2nd degree. Association analyses were conducted within each study, followed by inverse variance weighted fixed effects meta-analysis in METAL100. PheCodes were deemed statistically significant after Bonferroni correction for the number of PheCodes analyzed (P < 0.05/538 = 9.3 × 10−5).
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
The GWAS summary data generated in this study, including pooled, African, European, and sensitivity, have been deposited in the NHGRI-EBI Catalog of human genome-wide association studies (GWAS Catalog) database under accession codes GCST90502911 to GCST90502914 (https://www.ebi.ac.uk/gwas/downloads/summary-statistics). The raw TOPMed Program individual-level data are protected due to data privacy laws, but de-identified versions are available through Google and AWS cloud services following NIH dbGap approval. Details on gaining accessing are found on the TOPMed website (see https://topmed.nhlbi.nih.gov/topmed-data-access-scientific-community and https://topmed.nhlbi.nih.gov/topmed-whole-genome-sequencing-methods-freeze-8#access-to-sequence-data). In addition to raw data and full GWAS summary statistics provided through the referenced repositories, the summary statistics on the study population used in this study, along with summary results for top findings are provided in the Supplementary Data files.
Code availability
All protocols used for variant calling and quality control for data used in this study are described at: https://topmed.nhlbi.nih.gov/topmed-whole-genome-sequencing-methods-freeze-8. Links to relevant code, including GitHub repositories, are also provided on the same site. All GWAS analyses were performed using GENESIS, a Bioconductor package, on the TOPMed Analysis Commons. Details of the GENESIS app used on the Analysis Commons, along with underlying package information and code, are provided at: https://github.com/AnalysisCommons/genesis_wdl. All scripts used for running analyses on the TOPMed Analysis Commons using the DNAnexus Platform are available at: https://github.com/Justice-Genetics-Lab/TOPMed-WGS-BMI-GWAS/tree/main and https://doi.org/10.5281/zenodo.14708351.
References
Collaborators, G. B. D. O. et al. Health effects of overweight and obesity in 195 countries over 25 years. N. Engl. J. Med. 377, 13–27 (2017).
Chong, B. et al. Trends and predictions of malnutrition and obesity in 204 countries and territories: an analysis of the Global Burden of Disease Study 2019. EClinicalMedicine 57, 101850 (2023).
Fatumo, S. et al. A roadmap to increase diversity in genomic studies. Nat. Med. 28, 243–250 (2022).
Loos, R. J. F. & Yeo, G. S. H. The genetics of obesity: from discovery to biology. Nat. Rev. Genet. 23, 120–133 (2022).
Locke, A. E. et al. Genetic studies of body mass index yield new insights for obesity biology. Nature 518, 197–206 (2015).
Speakman, J. R., Loos, R. J. F., O’Rahilly, S., Hirschhorn, J. N. & Allison, D. B. GWAS for BMI: a treasure trove of fundamental insights into the genetic basis of obesity. Int. J. Obes. 42, 1524–1531 (2018).
Wainschtein, P. et al. Assessing the contribution of rare variants to complex trait heritability from whole-genome sequence data. Nat. Genet. 54, 263–273 (2022).
Taliun, D. et al. Sequencing of 53,831 diverse genomes from the NHLBI TOPMed Program. Nature 590, 290–299 (2021).
Abel, H. J. et al. Mapping and characterization of structural variation in 17,795 human genomes. Nature 583, 83–89 (2020).
Huang, L. et al. TOP-LD: A tool to explore linkage disequilibrium with TOPMed whole-genome sequence data. Am. J. Hum. Genet. 109, 1175–1181 (2022).
Tachmazidou, I. et al. Whole-genome sequencing coupled to imputation discovers genetic signals for anthropometric traits. Am. J. Hum. Genet. 100, 865–884 (2017).
McLaren, W. et al. The Ensembl variant effect predictor. Genome Biol. 17, 122 (2016).
Breeze, C. E. et al. FORGEdb: a tool for identifying candidate functional variants and uncovering target genes and mechanisms for complex diseases. Genome Biol. 25, 3 (2024).
ENCODE_Project_Consortium. An integrated encyclopedia of DNA elements in the human genome. Nature 489, 57–74 (2012).
Consortium, E. P. et al. Expanded encyclopaedias of DNA elements in the human and mouse genomes. Nature 583, 699–710 (2020).
Kircher, M. et al. A general framework for estimating the relative pathogenicity of human genetic variants. Nat. Genet. 46, 310–315 (2014).
Sullivan, P. F. et al. Leveraging base-pair mammalian constraint to understand genetic variation and human disease. Science 380, eabn2937 (2023).
Wen, W. et al. Meta-analysis identifies common variants associated with body mass index in east Asians. Nat. Genet. 44, 307–311 (2012).
Berndt, S. I. et al. Genome-wide meta-analysis identifies 11 new loci for anthropometric traits and provides insights into genetic architecture. Nat. Genet. 45, 501–512 (2013).
Gong, J. et al. Fine Mapping and Identification of BMI Loci in African Americans. Am. J. Hum. Genet. 93, 661–671 (2013).
Wen, W. et al. Meta-analysis of genome-wide association studies in East Asian-ancestry populations identifies four new loci for body mass index. Hum. Mol. Genet. 23, 5492–5504 (2014).
Winkler, T. W. et al. The influence of age and sex on genetic associations with adult body size and shape: a large-scale genome-wide interaction study. PLoS Genet. 11, e1005378 (2015).
Fernandez-Rhodes, L. et al. Trans-ethnic fine-mapping of genetic loci for body mass index in the diverse ancestral populations of the Population Architecture using Genomics and Epidemiology (PAGE) Study reveals evidence for multiple signals at established loci. Hum. Genet. 136, 771–800 (2017).
Ng, M. C. Y. et al. Discovery and fine-mapping of adiposity loci using high density imputation of genome-wide association studies in individuals of African ancestry: African Ancestry Anthropometry Genetics Consortium. PLoS Genet. 13, e1006719 (2017).
Justice, A. E. et al. Genome-wide meta-analysis of 241,258 adults accounting for smoking behaviour identifies novel loci for obesity traits. Nat. Commun. 8, 14977 (2017).
Akiyama, M. et al. Genome-wide association study identifies 112 new loci for body mass index in the Japanese population. Nat. Genet. 49, 1458–1467 (2017).
Turcot, V. et al. Protein-altering variants associated with body mass index implicate pathways that control energy intake and expenditure in obesity. Nat. Genet. 50, 26–41 (2018).
Gong, J. et al. Trans-ethnic analysis of metabochip data identifies two new loci associated with BMI. Int. J. Obes. 42, 384–390 (2018).
Jorgenson, E. et al. Genetic variation in the SIM1 locus is associated with erectile dysfunction. Proc. Natl Acad. Sci. USA 115, 11018–11023 (2018).
Yengo, L. et al. Meta-analysis of genome-wide association studies for height and body mass index in approximately 700000 individuals of European ancestry. Hum. Mol. Genet. 27, 3641–3649 (2018).
Kichaev, G. et al. Leveraging polygenic functional enrichment to improve GWAS power. Am. J. Hum. Genet. 104, 65–75 (2019).
Wojcik, G. L. et al. Genetic analyses of diverse populations improves discovery for complex traits. Nature 570, 514–518 (2019).
Fernandez-Rhodes, L. et al. Ancestral diversity improves discovery and fine-mapping of genetic loci for anthropometric traits-The Hispanic/Latino Anthropometry Consortium. HGG Adv. 3, 100099 (2022).
Akbari, P. et al. Sequencing of 640,000 exomes identifies GPR75 variants associated with protection from obesity. Science 373, eabf8683 (2021).
Kichaev, G. et al. Integrating functional data to prioritize causal variants in statistical fine-mapping studies. PLoS Genet. 10, e1004722 (2014).
Consortium, G. T. The Genotype-Tissue Expression (GTEx) project. Nat. Genet. 45, 580–585 (2013).
Laber, S. et al. Linking the FTO obesity rs1421085 variant circuitry to cellular, metabolic, and organismal phenotypes in vivo. Sci. Adv. 7, eabg0108 (2021).
Vamos, A. et al. Human abdominal subcutaneous-derived active beige adipocytes carrying FTO rs1421085 obesity-risk alleles exert lower thermogenic capacity. Front. Cell Dev. Biol. 11, 1155673 (2023).
Toth, B. B. et al. FTO intronic SNP strongly influences human neck adipocyte browning determined by tissue and PPARgamma specific regulation: a transcriptome analysis. Cells 9, 987 (2020).
Claussnitzer, M. et al. FTO obesity variant circuitry and adipocyte browning in humans. N. Engl. J. Med. 373, 895–907 (2015).
Maurano, M. T. et al. Large-scale identification of sequence variants influencing human transcription factor occupancy in vivo. Nat. Genet. 47, 1393–1401 (2015).
Jehan, S. et al. Obstructive sleep apnea and obesity: implications for public health. Sleep Med. Disord. 1, 00019 (2017).
Huang, J. et al. Genomics and phenomics of body mass index reveals a complex disease network. Nat. Commun. 13, 7973 (2022).
Cahoon, J. L. et al. Imputation accuracy across global human populations. Am. J. Hum. Genet. 111, 979–989 (2024).
Hubel, C. et al. Genomics of body fat percentage may contribute to sex bias in anorexia nervosa. Am. J. Med. Genet. B Neuropsychiatr. Genet. 180, 428–438 (2019).
Sakaue, S. et al. A cross-population atlas of genetic associations for 220 human phenotypes. Nat. Genet. 53, 1415–1424 (2021).
Vergne, I. & Deretic, V. The role of PI3P phosphatases in the regulation of autophagy. FEBS Lett. 584, 1313–1318 (2010).
Willer, C. J. et al. Discovery and refinement of loci associated with lipid levels. Nat. Genet. 45, 1274–1283 (2013).
Ghanbari, M. et al. A genetic variant in the seed region of miR-4513 shows pleiotropic effects on lipid and glucose homeostasis, blood pressure, and coronary artery disease. Hum. Mutat. 35, 1524–1531 (2014).
Geng, Z. et al. RNA-Seq analysis of obese Pdha1(fl/fl)Lyz2-Cre mice induced by a high-fat diet. Exp. Anim. 72, 112–122 (2022).
Brickner, J. R. et al. A ubiquitin-dependent signalling axis specific for ALKBH-mediated DNA dealkylation repair. Nature 551, 389–393 (2017).
Juszkiewicz, S., Speldewinde, S. H., Wan, L., Svejstrup, J. Q. & Hegde, R. S. The ASC-1 complex disassembles collided ribosomes. Mol. Cell 79, 603–614 e608 (2020).
Hashimoto, S., Sugiyama, T., Yamazaki, R., Nobuta, R. & Inada, T. Identification of a novel trigger complex that facilitates ribosome-associated quality control in mammalian cells. Sci. Rep. 10, 3422 (2020).
Narita, M. et al. A distinct mammalian disome collision interface harbors K63-linked polyubiquitination of uS10 to trigger hRQT-mediated subunit dissociation. Nat. Commun. 13, 6411 (2022).
Patitucci, C. et al. Mtfp1 ablation enhances mitochondrial respiration and protects against hepatic steatosis. Nat. Commun. 14, 8474 (2023).
Han, J. et al. Transcriptome profiling of developing ovine fat tail tissue reveals an important role for MTFP1 in regulation of adipogenesis. Front. Cell Dev. Biol. 10, 839731 (2022).
Landrum, M. J. et al. ClinVar: improving access to variant interpretations and supporting evidence. Nucleic Acids Res. 46, D1062–D1067 (2018).
National Center for Biotechnology Information. ClinVar; [VCV001233562.6], https://www.ncbi.nlm.nih.gov/clinvar/variation/VCV001233562.6 [accessed Jan. 10, 2023]
Kanai, M. et al. Insights from complex trait fine-mapping across diverse populations. Preprint at medRxiv https://doi.org/10.1101/2021.09.03.21262975 (2021).
Spain, S. L. & Barrett, J. C. Strategies for fine-mapping complex traits. Hum. Mol. Genet. 24, R111–R119 (2015).
Kichaev, G. & Pasaniuc, B. Leveraging functional-annotation data in trans-ethnic fine-mapping studies. Am. J. Hum. Genet. 97, 260–271 (2015).
Yuan, K. et al. Fine-mapping across diverse ancestries drives the discovery of putative causal variants underlying human complex traits and diseases. Nat. Genet. 56, 1841–1850 (2024).
Benner, C. et al. FINEMAP: efficient variable selection using summary data from genome-wide association studies. Bioinformatics 32, 1493–1501 (2016).
Wang, G., Sarkar, A., Carbonetto, P. & Stephens, M. A simple new approach to variable selection in regression, with application to genetic fine mapping. J. R. Stat. Soc. Ser. B Stat. Methodol. 82, 1273–1300 (2020).
Hotta, K. et al. Association between obesity and polymorphisms in SEC16B, TMEM18, GNPDA2, BDNF, FAIM2 and MC4R in a Japanese population. J. Hum. Genet. 54, 727–731 (2009).
Speliotes, E. K. et al. Association analyses of 249,796 individuals reveal 18 new loci associated with body mass index. Nat. Genet 42, 937–948 (2010).
Conomos, M. P., Reiner, A. P., Weir, B. S. & Thornton, T. A. Model-free Estimation of Recent Genetic Relatedness. Am. J. Hum. Genet. 98, 127–148 (2016).
Khan, A. T. et al. Recommendations on the use and reporting of race, ethnicity, and ancestry in genetic research: experiences from the NHLBI TOPMed program. Cell Genom. 2, 100155 (2022).
Fang, H. et al. Harmonizing genetic ancestry and self-identified race/ethnicity in genome-wide association studies. Am. J. Hum. Genet. 105, 763–772 (2019).
Regier, A. A. et al. Functional equivalence of genome sequencing analysis pipelines enables harmonized variant calling across human genetics projects. Nat. Commun. 9, 4038 (2018).
Jun, G., Wing, M. K., Abecasis, G. R. & Kang, H. M. An efficient and scalable analysis framework for variant extraction and refinement from population-scale DNA sequence data. Genome Res. 25, 918–925 (2015).
Gogarten, S. M. et al. Genetic association testing using the GENESIS R/Bioconductor package. Bioinformatics 35, 5346–5348 (2019).
Brody, J. A. et al. Analysis commons, a team approach to discovery in a big-data environment for genetic epidemiology. Nat. Genet. 49, 1560–1563 (2017).
Sofer, T. et al. A fully adjusted two-stage procedure for rank-normalization in genetic association studies. Genet. Epidemiol. 43, 263–275 (2019).
Conomos, M. P., Miller, M. B. & Thornton, T. A. Robust inference of population structure for ancestry prediction and correction of stratification in the presence of relatedness. Genet. Epidemiol. 39, 276–293 (2015).
Conomos, M. P. et al. Genetic diversity and association studies in US Hispanic/Latino populations: applications in the Hispanic community health study/study of Latinos. Am. J. Hum. Genet. 98, 165–184 (2016).
Lin, D. Y. A simple and accurate method to determine genomewide significance for association tests in sequencing studies. Genet. Epidemiol. 43, 365–372 (2019).
Abul-Husn, N. S. et al. Implementing genomic screening in diverse populations. Genome Med. 13, 17 (2021).
Gaziano, J. M. et al. Million veteran program: a mega-biobank to study genetic influences on health and disease. J. Clin. Epidemiol. 70, 214–223 (2016).
Kolonel, L. N. et al. A multiethnic cohort in Hawaii and Los Angeles: baseline characteristics. Am. J. Epidemiol. 151, 346–357 (2000).
Carey, D. J. et al. The Geisinger MyCode community health initiative: an electronic health record-linked biobank for precision medicine research. Genet. Med. 18, 906–913 (2016).
Howard, V. J. et al. The reasons for geographic and racial differences in stroke study: objectives and design. Neuroepidemiology 25, 135–143 (2005).
Sudlow, C. et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 12, e1001779 (2015).
Bycroft, C. et al. The UK Biobank resource with deep phenotyping and genomic data. Nature 562, 203–209 (2018).
Stata Statistical Software: Release 15.1 (StataCorp, 2017).
Graff, M. et al. Genome-wide physical activity interactions in adiposity - a meta-analysis of 200,452 adults. PLoS Genet. 13, e1006528 (2017).
Machiela, M. J. & Chanock, S. J. LDlink: a web-based application for exploring population-specific haplotype structure and linking correlated alleles of possible functional variants. Bioinformatics 31, 3555–3557 (2015).
Karczewski, K. J. et al. The mutational constraint spectrum quantified from variation in 141,456 humans. Nature 581, 434–443 (2020).
Consortium, F. et al. A promoter-level mammalian expression atlas. Nature 507, 462–470 (2014).
Fishilevich, S. et al. GeneHancer: genome-wide integration of enhancers and target genes in GeneCards. Database 2017 https://doi.org/10.1093/database/bax028 (2017).
Chen, H. et al. Efficient variant set mixed model association tests for continuous and binary traits in large-scale whole-genome sequencing studies. Am. J. Hum. Genet. 104, 260–274 (2019).
Liu, X., Wu, C., Li, C. & Boerwinkle, E. dbNSFP v3.0: a one-stop database of functional predictions and annotations for human nonsynonymous and splice-site SNVs. Hum. Mutat. 37, 235–241 (2016).
Chen, W. et al. Fine mapping causal variants with an approximate Bayesian method using marginal test statistics. Genetics 200, 719–736 (2015).
LaPierre, N. et al. Identifying causal variants by fine mapping across multiple studies. PLoS Genet. 17, e1009733 (2021).
Greenbaum, J. & Deng, H. W. A statistical approach to fine mapping for the identification of potential causal variants related to bone mineral density. J. Bone Min. Res. 32, 1651–1658 (2017).
Conrad, D. F. et al. A worldwide survey of haplotype variation and linkage disequilibrium in the human genome. Nat. Genet. 38, 1251–1260 (2006).
Abul-Husn, N. S. & Kenny, E. E. Personalized medicine and the power of electronic health records. Cell 177, 58–69 (2019).
Wu, P. et al. Mapping ICD-10 and ICD-10-CM codes to phecodes: workflow development and initial evaluation. JMIR Med. Inf. 7, e14325 (2019).
Carroll, R. J., Bastarache, L. & Denny, J. C. R PheWAS: data analysis and plotting tools for phenome-wide association studies in the R environment. Bioinformatics 30, 2375–2376 (2014).
Willer, C. J., Li, Y. & Abecasis, G. R. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics 26, 2190–2191 (2010).
Acknowledgements
Individual researchers received funding for their work, including National Institute of Health (NIH) grants: R01 DK122503 (C.T.L., K.E.N., A.E.J., G.C., N.S.J.), T32 HL007055 (H.M.H.), T32 HL129982 (H.M.H.), R01HL142825 (H.M.H.), I01-BX003362 (K.C.), U01 HL120393 (M.P.C., B.D.He., D.J.), R01 HL68959 (M.E.G., A.E.A., M.J.T.), U01 HL072507 (J.E.H., J.H.), K08 HL136928 (B.D.Ho.), R01HL-120393 (D.J.), R01 HL119443 (S.L.R.K.), R01 HL055673-18S1 (S.L.R.K.), R01 HL92301 (N.D.P.), R01 HL67348 (N.D.P.), R01 NS058700 (N.D.P.), R0 AR48797 (N.D.P.), R01 DK071891 (N.D.P.), R01 AG058921 (N.D.P.), F32 HL085989 (N.D.P.), U01 HL089897 (E.A.R.), U01 HL089856 (E.A.R.), R01 HL093093 (D.E.W.), R01 HL133040 (D.E.W.), I01 BX003340 (P.W.F.W.), I01 BX004821 (P.W.F.W.), U01 HL072524 (D.K.A.), R01 HL104135-04S1 (D.K.A.), U01 HL054472 (D.K.A.), U01 HL054473 (D.K.A.), U01 HL054495 (D.K.A.), U01 HL054509 (D.K.A.), R01 HL055673 with supplement -18S1 (D.K.A.), R01 HL104608 (K.C.B.), R01 AI132476 (K.C.B.), R01 AI114555 (K.C.B.), R01 HL104608-S1 (K.C.B.), P20 GM109036 (J.H.), KL2 TR002490 (L.M.R.), T32 HL129982 (L.M.R.), P01 HL132825 (S.T.W.), R35 CA197449 (X.Lin), P01 CA134294 (X.Lin), U19 CA203654 (X.Lin), R01 HL113338 (X.Lin), U01 HG009088 (X.Lin), R01 HL142302 (R.J.F.L.), R01 DK124097 (R.J.F.L.), R01 DK110113 (R.J.F.L.), R01 DK107786 (R.J.F.L.), R01 DK075787 (R.J.F.L.), NNF23SA0084103 (R.J.F.L.), NNF18CC0034900 (R.J.F.L.), NNF20OC0059313 (R.J.F.L.), X01 HL134588 (R.J.F.L.), R01 HG010297 (K.E.N.), U01 HG007416 (K.E.N.), R01 DK135938 (N.S.J.), R01 HL105756 (B.M.P.), NHLBI TOPMed Fellowship (X.Li), P30 CA008748 (M.D.); contracts HHSN268201800001I (M.P.C., B.D.He., D.J.), HHSN268201500014C (S.L.R.K.); American Diabetes Association (ADA) Grant 1-19-PDF-045 (H.M.H.); the General Clinical Research Center of the Wake Forest University School of Medicine, M01 RR07122 (N.D.P.); and a pilot grant from the Claude Pepper Older Americans Independence Center of Wake Forest University Health Sciences, P60 AG10484 (N.D.P.).
Author information
Authors and Affiliations
Consortia
Contributions
Conducted analyses or contributed to figures and tables: X.Z., J.A.B., M.G., N.C., Zh.W., K.F., G.C., N.S.J., Q.H., Ze.W., M.M., S.G., and A.E.J. Supervised analyses: Y.V.S., L.A.C., L.A.L., C.T.L., R.J.F.L., K.E.N., and A.E.J. Contributed to the design of the current study: X.Z., J.A.B., M.G., H.M.H., N.C., H.X., L.A.C., L.A.L., C.T.L., R.J.F.L., K.E.N., and A.E.J. Contributed to the conception or design of the TOPMed program and its operations (including organization and policies of TOPMed – e.g., exec committee, working group conveners, NHLBI staff, etc.): N.C., Zh.W., D.L.D., B.D.He., J.E.H., S.L.R.K., T.N.K., J.S.P., E.A.R., R.S.V., J.W., D.K.A., K.C.B., J.H., S.R.H., B.M.P., L.M.R., S.S.R., J.I.R., N.L.S., K.D.T., L.A.C., C.T.L., R.J.F.L., K.E.N., and A.E.J. Provided phenotypic data and/or biosamples: M.G., N.C., M.A.A., L.F.B., M.P.B., S.C., D.L.D., R.D., X.G., C.Ha., B.H., J.E.H., Y.H., R.D.J., S.L.R.K., E.M.L., L.L.M., T.N., N.D.P., M.H.P., E.A.R., S.M.R., D.M.D., R.S.V., D.E.W., J.W., L.R.W., L.R.Y., Z.T.Y., D.K.A., J.B., E.G.B., A.P.C., D.I.C., J.E.C., M.F., J.H., S.R.H., C.K., R.M., B.M.P., L.M.R., A.P.R., S.S.R., J.I.R., M.B.S., N.L.S., N.H., L.A.C., L.A.L., R.J.F.L., K.E.N., and A.E.J. Acquired WGS and/or other omics data: L.F.B., D.L.D., R.D., J.E.H., S.L.R.K., J.A.S., R.S.V., J.W., W.Z., D.K.A., J.B., E.G.B., J.E.C., M.F., J.H., S.R.H., C.K., B.M.P., A.P.R., S.S.R., J.I.R., N.L.S., K.D.T., L.A.C., R.J.F.L., K.E.N., and A.E.J. Created software, processed, and/or analyzed WGS or other study data for data summaries in this paper: X.Z., J.A.B., M.G., H.M.H., N.C., J.C.B., J.G.B., E.J.B., M.P.C., B.D.He., C.L., C.P.M., K.L.W., and A.E.J. Drafted the manuscript and revised according to co-author suggestions: X.Z., J.A.B., M.G., H.M.H., N.C., K.E.N., and A.E.J. Critically reviewed the manuscript, suggested revisions as needed, and approved the final version: X.Z., J.A.B., M.G., H.M.H., N.C., H.X., Zh.W., K.F., G.C., N.S.J., M.M., S.G., X.Li., Z.L., M.A.A., D.M.B., L.F.B., J.C.B., M.P.B., D.W.B., J.G.B., E.J.B., C.S.C., K.C., S.C., Y.C., L.C., M.P.C., D.L.D., M.D., R.D., C.E., A.E.F., B.I.F., M.E.G., X.G., C.Ha., B.D.He., B.H., J.E.H., Y.H., B.D.Ho., D.H., Q.H., C.Hw., R.D.J., D.J., R.R.K., S.L.R.K., T.N.K., E.M.L., M.L., C.L., L.L.M., M.N.M., C.P.M., A.C.M., T.N., J.O., C.J.O., N.D.P., J.S.P., J.A.P., U.P., M.H.P., D.C.R., E.A.R., S.M.R., D.M.D., J.R., C.M.S., J.A.S., H.K.T., R.S.V., Ze.W., D.E.W., D.I.C., J.W., K.L.W., L.R.W., P.W.F.W., L.R.Y., Z.T.Y., W.Z., S.Z., D.K.A., A.E.A., K.C.B., J.B., E.B., E.G.B., A.P.C., Y.I.C., J.E.C., M.F., V.R.G., J.H., S.R.H., L.H., M.R.I., C.K., R.M., B.D.M., M.N., B.M.P., L.M.R., A.P.R., S.S.R., J.I.R., M.B.S., N.L.S., K.D.T., M.J.T., S.T.W., Y.Z., N.H., Y.V.S., X.Lin, L.A.C., L.A.L., C.T.L., R.J.F.L., K.E.N., and A.E.J.
Corresponding authors
Ethics declarations
Competing interests
B.D.Ho. receives grant support from Bayer and has received an honorarium from AstraZeneca for an educational lecture. B.M.Ps. serve on the TOPMed Steering Committee. C.J.O. is employed by Novartis Institute of Biomedical Research, Cambridge, MA. D.L.D. received grants from Bayer and honoraria from Novartis. K.C.B. is an employee of Tempus. L.M.R. and S.S.R. are consultants for the TOPMed Administrative Coordinating Center (through Westat). U.P. was a consultant with AbbVie, and her husband is holding individual stocks for the following companies: BioNTech SE – ADR, Amazon, CureVac BV, NanoString Technologies, Google/Alphabet Inc Class C, NVIDIA Corp, Microsoft Corp. XLin is a consultant of AbbVie Pharmaceuticals and Verily Life Sciences. The remaining authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Zhang, X., Brody, J.A., Graff, M. et al. Whole genome sequencing analysis of body mass index identifies novel African ancestry-specific risk allele. Nat Commun 16, 3470 (2025). https://doi.org/10.1038/s41467-025-58420-2
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-025-58420-2
