Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Evidence of Polygenic Adaptation in the Systems Genetics of Anthropometric Traits

  • Renato Polimanti ,

    renato.polimanti@yale.edu

    Affiliations Department of Psychiatry, Yale University School of Medicine, West Haven, Connecticut, United States of America, VA CT Healthcare Center, West Haven, Connecticut, United States of America

  • Bao Zhu Yang,

    Affiliations Department of Psychiatry, Yale University School of Medicine, West Haven, Connecticut, United States of America, VA CT Healthcare Center, West Haven, Connecticut, United States of America

  • Hongyu Zhao,

    Affiliations Department of Biostatistics, Yale University School of Public Health, New Haven, Connecticut, United States of America, Department of Genetics, Yale University School of Medicine, New Haven, Connecticut, United States of America

  • Joel Gelernter

    Affiliations Department of Psychiatry, Yale University School of Medicine, West Haven, Connecticut, United States of America, VA CT Healthcare Center, West Haven, Connecticut, United States of America, Department of Genetics, Yale University School of Medicine, New Haven, Connecticut, United States of America, Department of Neurobiology, Yale University School of Medicine, New Haven, Connecticut, United States of America

Evidence of Polygenic Adaptation in the Systems Genetics of Anthropometric Traits

  • Renato Polimanti, 
  • Bao Zhu Yang, 
  • Hongyu Zhao, 
  • Joel Gelernter
PLOS
x

Abstract

Many signals of natural selection have been identified in the human genome. However, except for some single-locus mechanisms, most molecular processes generating these adaptation signals are still unknown. We developed an approach that integrates datasets related to genome-wide association studies (GWAS) with information about systems biology and genetic signatures of natural selection to identify evidence of polygenic adaptation. Specifically, we focused on five anthropometric measurements: body mass index (BMI), height, waist-to-hip ratio adjusted for BMI (WHR), and waist circumference adjusted for BMI (WC), and sex differences for WHR and WC. We performed an enrichment analysis for signals of natural selection in protein interaction networks associated with anthropometric traits in European populations. The adaptation signals-enriched gene networks associated highlighted epistatic interactions in the context of polygenic selection for the investigated traits. These polygenic mechanisms indicated intriguing selective mechanisms related to the anthropometric traits: adult locomotory behavior for BMI, infection resistance for height, interplay between lipid transport and immune systems for WHR, and female-specific polygenic adaptation for WHR and WC. In conclusion, we observed evidence of polygenic adaptation in the context of systems genetics of anthropometric traits that indicates polygenic mechanisms related to the natural selection in European populations.

Introduction

Numerous studies have investigated the human genome to understand how human species adapted to the very different environments worldwide [1, 2]. The explosion of high-throughput technologies and informatics permitted development of methods based on genome-wide data to detect signatures of natural selection in the human genome [3]. Accordingly, reliable evidence of genetic signals of selection across the genome has been identified regarding human adaptations to infections [4], ultraviolet radiation [5], diet [6], and high altitude [7]. However, some investigators have argued that the identified single-locus signals of selection represent only a small part of the process of shaping adaptation-related human genomic variation [8, 9]; and that most adaptation events in natural populations occur via polygenic mechanisms, in accordance with quantitative genetics models. Genome-wide association studies (GWAS) have widely demonstrated that complex traits are substantially attributable to many independent loci with small effect sizes [10] and genetic adaptation can affect GWAS outcomes [11]. A broader view of these findings in the context of natural selection is most consistent with the view that selection of phenotypic traits in response to environmental pressures probably generally occurs via polygenic adaptation, rather than sudden large-effect mutations at individual key loci. Although these data strongly encourage exploration of polygenic mechanisms, this is still a field in its infancy. In 2012, Turchin and colleagues detected signature of widespread selection in height-associated loci [12]. Specifically, they observed that frequencies of alleles associated with increased height are systematically elevated in Northern Europeans compared with Southern Europeans. In 2013, Daub and colleagues used a different approach based on FST probabilities, pathway-enrichment analysis, and long-distance genotypic linkage disequilibrium (LD) to investigate polygenic adaptation in human genome [13]. They showed that pathogen-human interactions have produced widespread and coordinated genomic responses, indicating that adaptation to pathogens is a good example of polygenic selection. Berg and Coop investigated polygenic adaptation among the loci associated with different complex traits (i.e., height, skin pigmentation, body mass index (BMI), type 2 diabetes, Crohn’s disease, and ulcerative colitis) by GWAS; they observed a number of putative signals of local adaptation [14]. These investigations provided evidence regarding polygenic adaptation and methods useful to investigate polygenic adaptation.

Compared to the previous studies which used information about GWAS significant outcomes or pre-defined molecular pathways, we employed GWAS datasets (i.e., data about millions of variants across the human genome) together with information about systems biology and genetic signatures of natural selection. To test our approach, we focused this investigation on different sub-phenotypes (i.e., distribution, phenotypic variability, extreme phenotype differences, and sex-differences) related different anthropometric traits (BMI, height, waist-to-hip ratio adjusted for BMI (WHR), and waist circumference adjusted for BMI (WC)), since anthropometric traits have been affected by numerous different environmental pressures in the process of human evolution and adaptation, including known adaptation mechanisms that link anthropometric traits with climate [15], diet [16], and fertility [17]. Furthermore, since the polygenic inheritance of anthropometric traits seems not attributable to lower-frequency variants [18], these traits are good candidates to be investigated by our approach that focused attention on polygenic inheritance from common variants. In our investigation, we performed iHS (integrated Haplotype score) enrichment analysis of protein interaction networks associated with anthropometric traits, tested the epistatic interactions among loci included in the iHS-enriched protein networks, and verified the enrichments of the interactive loci for gene ontologies and known molecular pathways. Since haplotype structure-based statistics can identify selective sweeps under a number of different selection scenarios (i.e., complete, incomplete hard, and soft sweeps) [19], our iHS-based strategy is capable of detecting polygenic adaptation related to several different selection mechanisms. Our findings indicated evidence of polygenic adaptation in the context of systems genetics of anthropometric traits that are related to different biological mechanisms.

Materials and Methods

GWAS summary statistics

We considered the large-scale meta-analysis of GWAS performed by GIANT (Genetic Investigation of ANthropometric Traits) consortium for BMI [2022], height [2123], WHR [22, 24, 25], and WC [25]. GIANT data were downloaded from http://www.broadinstitute.org/collaboration/giant/index.php/GIANT_consortium_data_files in August 2014. For each anthropometric trait, we investigated GWAS data related to the analyses based on different sub-phenotypes: distribution, phenotypic variability, and extreme phenotype differences (S1 Table). The GIANT investigators observed sex differences in the genetics of WC and WHR [25], so we analyzed male- and female-specific GWAS data for these two traits separately. Regarding WC and WHR analysis, we focused on BMI-adjusted traits since these adjusted traits provide information about body shape and fat accumulation distribution. Since no summary statistics are publically available about large GWAS of anthropometric traits in non-European populations, we restricted our analysis to the GIANT data and the genetic variation within European ancestry. Our recent study demonstrated that GWAS findings depends to the ancestry genomic background [11]. Accordingly, summary statistics of GWAS performed on European populations cannot be used to investigate other human groups.

Integrated haplotype score screening in protein interaction networks

We used information about iHS–a statistic based on the differential levels of LD able to detect evidence of recent positive selection [26]–to focus our analysis on loci which are known to have suggestive signals of natural selection. We chose this specific selection metric because this method is appropriate to detect non-fixed selected alleles, like those expected in a polygenic adaptation scenario [9], in a specific population. A recent study confirmed that haplotype structure-based statistics can identify selective sweeps under several different selection scenarios (i.e., complete, incomplete hard, and soft sweeps) [19].

Since GIANT performed their GWAS on European ancestry populations, we used the iHS data based on the HapMap Phase 2 CEU (Utah residents with Northern and Western European ancestry from the CEPH collection) population available in Haplotter database [26]. Previous studies used |iHS| > 1.5 as suggestive evidence for natural selection [2729], in agreement with a recent simulation analysis that indicated |iHS| > 2.0 is too stringent a threshold [30]. Accordingly, we selected from GIANT GWAS datasets those variants with |iHS| > 1.5 (S1 Table). Considering iHS selection criterion, we performed a gene-based association analysis using VEGAS software [31]. The HapMap CEU reference panel was used to correct for LD patterns. We used the summary statistics of gene-based association analysis to perform an association analysis based on protein-protein interactions (PPI) using the R package dmGWAS [32]. We defined PPIs of all genes with gene-based association using the Protein Interaction Network Analysis (PINA) platform v2.0 [33] that includes information about physical and genetic PPIs. For each investigated trait, we further considered analysis for the associated gene network defined by the top-10 PPI modules (S2 Table). We developed this multi-step approach based on iHS screening and gene- and PPI-based analyses to apply the iHS enrichment analysis on protein interaction networks, instead of a continuous genome region. This approach can improve the effectiveness of iHS screening to detect polygenic adaptation rather than single-locus signals.

Long-distance genotypic Linkage Disequilibrium

As previously performed by Daub and colleagues [13], we used long-distance genotypic LD to detect epistatic interactions among variants included in the protein networks enriched for selection signals. For each trait, we calculated the long-distance genotypic LD among the variants located in the genes included in the top-10 PPI modules using Genepop 4.2 software [34]. To reduce the number of variants carried forward to investigate in the next steps, we pruned the variants for LD (r2 > 0.8) using PLINK software [35]. Both data pruning and genotypic LD analysis were performed using HapMap Phase 3 CEU population [36]. Long-distance LD pairs were defined as pairs of variants either (a) mapped to different chromosomes, or (b) on the same chromosome separated by ≥ 10 Mb. Furthermore, we excluded LD pairs where none of the variants is associated with the anthropometric trait at a nominal significance level. We used the R package “qvalue” to estimate q-values for controlling the false discovery rate (FDR), and considered q-values < 0.05 as significant.

Term Enrichment analysis

Considering the genes with significant evidence of epistatic interactions (long-distance genotypic LD with FDR q<0.05), we generated a gene set for each investigated anthropometric trait potentially and analyzed the enrichment for different categorical data (e.g., gene ontology and molecular pathway membership) using DAVID 6.7 software [37].

Results

Body Mass Index

BMI PPI-based analysis generated three gene networks associated with the investigated BMI-related sub-phenotypes (i.e., distribution, phenotypic variability, and extreme phenotype differences). The three BMI-related gene networks (S3 Table) respectively include 201 genes (BMI distribution; 1,013 pruned variants with |iHS| > 1.5), 93 genes (BMI phenotypic variability; 395 pruned variants with |iHS| > 1.5), and 79 genes (BMI extreme phenotype differences; 332 pruned variants with |iHS| > 1.5). Calculating the long-distance genotypic LD, we observed significant LD pairs (Fig 1; S4 Table): 20 in the BMI distribution gene network, three in the BMI phenotypic-variability gene network, and two in the BMI extreme-phenotypic-difference gene network. Considering the information about long-distance genotypic LD and PPIs, we generated a BMI gene set based on the putative signals of polygenic adaptation in the systems genetics of BMI (Fig 2). The enrichment analysis in this gene set indicated that it is significantly associated with adult locomotory behavior, learning, and various molecular regulatory mechanisms (Table 1).

thumbnail
Fig 1. Significant long-distance genotypic LD in BMI-associated gene network.

The BMI phenotypic analyses are highlighted with different colors: green (distribution), red (phenotypic variability), blue (extreme phenotypes), and black (overlapping between distribution and extreme-phenotype analyses).

https://doi.org/10.1371/journal.pone.0160654.g001

thumbnail
Fig 2. Gene set with evidences of polygenic adaptation in the systems genetics of BMI-related sub-phenotypes.

The intensity of the box shadings is proportional to the strength of the gene-based significant association. The types of line indicate the source of interaction evidence: solid line (BMI distribution analysis), dashed line (BMI extreme phenotype analysis), dotdash line (phenotypic variability analysis), and dotted line (PINA database).

https://doi.org/10.1371/journal.pone.0160654.g002

thumbnail
Table 1. Term enrichment analysis of gene sets with evidences of polygenic adaptation.

Fisher exact test p values adjusted for Bonferroni correction are reported.

https://doi.org/10.1371/journal.pone.0160654.t001

Height

Height PPI-based association analysis produced three gene networks for the investigated height-related sub-phenotypes (i.e., distribution, phenotypic variability, and extreme phenotype differences). Specifically, the height-related gene networks (S5 Table) respectively include 532 genes for height distribution (2,246 pruned variants with |iHS| > 1.5), 114 genes for phenotypic variability (447 pruned variants with |iHS| > 1.5), and 409 genes for extreme phenotype differences (1,692 pruned variants with |iHS| > 1.5). The analysis of long-distance genotypic LD identified numerous significant LD pairs (Fig 3; S6 Table): 68 in the height distribution gene network, two in the height phenotypic-variability gene network, and 25 in the height extreme-phenotypic-difference gene network. The gene sets made using the information about the long-distance genotypic LD and PPIs suggested numerous interacting processes in polygenic adaptation related to height (Fig 4). The enrichment analysis indicated that this height-related gene set is significantly associated with GWAS outcomes of AIDS-nonprogression phenotype, GWAS results for height, immune systems-related functions, and other molecular processes (Table 1).

thumbnail
Fig 3. Significant long-distance genotypic LD in Height-associated gene network.

The height phenotypic analyses are highlighted with different colors: green (distribution), red (phenotypic variability), and blue (extreme phenotypes).

https://doi.org/10.1371/journal.pone.0160654.g003

thumbnail
Fig 4. Gene set with evidences of polygenic adaptation in the systems genetics of Height-related sub-phenotypes.

The intensity of the box shadings is proportional to the strength of the gene-based significant association. The types of line indicate the source of interaction evidence: solid line (height distribution analysis), dashed line (height phenotypic variability analysis), dotdash line (height extreme phenotype analysis), and dotted line (PINA).

https://doi.org/10.1371/journal.pone.0160654.g004

Waist circumference

For WC, GWAS datasets were available only for sex differences (i.e., men and women). WC PPI-based association analysis generated a gene network for each investigated WC-related sub-phenotype (S7 Table). A total of 148 genes were included in the male WC gene network (715 pruned variants with |iHS| > 1.5), and 77 genes in the female WC gene network (328 pruned variants with |iHS| > 1.5). In these gene networks, we observed one and two significant long-distance genotypic LD pairs, respectively (Fig 5; S8 Table). Fig 6 reports the WC gene set based on genotypic LD analysis and PPI information. No significant enrichments were observed in the WC gene set.

thumbnail
Fig 5. Significant long-distance genotypic LD in WC-associated gene network.

The WC phenotypic analyses are highlighted with different colors: yellow (men) and purple (women).

https://doi.org/10.1371/journal.pone.0160654.g005

thumbnail
Fig 6. Gene set with evidences of polygenic adaptation in the systems genetics of WC-related sub-phenotypes.

The intensity of the box shadings is proportional to the strength of the gene-based significant association. The types of line indicate the source of interaction evidence: dashed line (women WC analysis) and dotted line (men WC analysis).

https://doi.org/10.1371/journal.pone.0160654.g006

Waist-to-hip ratio

In WHR analysis, we considered four sub-phenotypes: distribution and extreme phenotype differences, and males and females. PPI-based association analysis defined a gene network for each investigated WHR sub-phenotype (S9 Table). One hundred thirty-nine genes are included in the WHR-distribution gene network (702 pruned variants with |iHS| > 1.5), 138 in the WHR-extreme-phenotype-difference gene network (578 pruned variants with |iHS| > 1.5), 65 genes in the WHR-men gene network (257 pruned variants with |iHS| > 1.5), and 92 genes in the WHR-women gene network (406 pruned variants with |iHS| > 1.5). Among these WHR-related gene networks, we observed significant long-distance genotypic LD (Fig 7; S10 Table): two significant pairs in WHR distribution, five significant pairs in WHR-extreme-phenotype-difference, one significant pair in WHR-males, and four significant pairs in WHR-females. Fig 8 reports the WHR gene sets based on long-distance genotypic LD and PPIs. The enrichment analysis indicates a significant association for cellular structure-related terms in male WHR gene sets, and terms related to immune systems and lipid transportation in female WHR gene sets (Table 1).

thumbnail
Fig 7. Gene set with evidences of polygenic adaptation in the systems genetics of WHR-related sub-phenotypes.

The intensity of the box shadings is proportional to the strength of the gene-based significant association. The types of line indicate the source of interaction evidence: solid line (WHR distribution analysis), dashed line (WHR extreme phenotype analysis), dotdash line (WHR men analysis), double line (WHR women analysis), and dotted line (PINA).

https://doi.org/10.1371/journal.pone.0160654.g007

thumbnail
Fig 8. Significant long-distance genotypic LD in WHR-associated gene network.

The WHR phenotypic analyses are highlighted with different colors: green (distribution), blue (extreme phenotypes), yellow (men) and purple (women).

https://doi.org/10.1371/journal.pone.0160654.g008

Discussion

Our results highlight evidences of polygenic adaptation in the systems genetics of anthropometric traits, suggesting that polygenic adaptation is not uncommon in the systems genetics of complex traits, and confirming that anthropometric traits are influenced by several selective pressures related to biological mechanisms, such as behaviors and immune systems. Our approach demonstrated the usefulness of combining information from GWAS and systems biology in investigating polygenic mechanisms in human adaptation. Previous studies of polygenic adaptation obtained interesting results [1214], focusing on main genetic risk factors or specific pathways, and investigating other natural selection signatures in human genome (e.g., FST). In our study, we applied a different approach based on GWAS summary statistics, gene interactive network and iHS. Since polygenic mechanisms mostly involve multiple loci with small effects that are implicated in different molecular pathways [10], our approach appears, at least in these instances, to be effective in detecting polygenic selection in the systems genetics of complex traits. Thus we have demonstrated utility in combining GWAS of complex traits with information about systems biology and natural selection signatures. Furthermore, we also confirmed the recent analysis of Ferrer-Admetlla and colleagues [19] regarding haplotype structure-based statistics since our iHS-based analysis was able to detect the widespread signals of polygenic adaptation.

In the BMI analysis, an examination of the systems genetics of this trait resulted in a conclusion that there is evidence of polygenic adaptation in the systems genetics of this trait. Most significant epistatic interactions (i.e., significant long-distance genotypic LD) among variants with suggestive signatures of natural selection (i.e., |iHS| > 1.5) are related to the analysis of BMI distributions. This is almost certainly due to the fact that GWAS of BMI distribution identified more loci associated with this sub-phenotype than the other BMI-related sub-phenotypes, and therefore this comparison had better power. The enrichment analysis of the BMI gene sets related to polygenic adaptation indicated significant association with some gene ontologies. Among them, the most obviously relevant is related to adult locomotory behavior and learning. Adult locomotory behavior is defined as the specific movement from place to place of a fully developed and mature organism in response to external or internal stimuli. Animal experiments explored different molecular aspects involved in stimulus-response activities, highlighting the interplay between food intake and locomotory behavior to maintain energy homoeostasis [3840]. No previous study highlighted putative signals of natural selection in relation to genes involved in these mechanisms. Our data support an intriguing scenario where genes involved in adult locomotory behavior are under selective pressure, and this could affect the systems genetics of BMI. Furthermore, recent data indicated that some BMI-associated loci mechanisms may be present in specific population categories [41], suggesting additional pathogenic mechanisms that differ from those present in the general population.

In the height analysis, we identified a large gene set with evidence of polygenic adaptation. This is consistent with both the previous studies regarding natural selection related to height [12, 14] and the large number of genome-wide significant loci identified by GWAS of height [23, 42]. Our results provide additional information about the interplay between height and immune system. The gene set with evidence of polygenic adaptation related to height distribution is enriched in genes identified by GWAS of AIDS resistance [43], and genes encoding natural killer cell lectin-like receptor binding. Several studies demonstrated the relevant relationship between infection occurrences and stature, suggesting a relevant interplay between nutrition and infection [43, 44]. Furthermore, a recent study indicated high frequency of short stature in HIV-infected children, and poor adherence to antiretroviral treatment, severe immunosuppression, and therapy inefficacy are associated with severe short stature.[45] Accordingly, it is a reasonable hypothesis that greater infection susceptibility is one reason that individuals might fail to attain what would otherwise be their genetically-determined greatest height; and that adaptation processes related to infection resistance could also play a role in the genetics of height. Moreover, one of the previous studies about polygenic adaptation found widespread and coordinated genomic responses related to the adaptation to pathogens [13], consistent with our findings about polygenic selection in immune system-related genes.

WC analysis identified little evidence of polygenic adaptation, and what was observed was mainly in females. No significant term enrichments were observed. However, among the genes included in the female WC gene set, three may be interesting: CRHR1, NOS3, and FARS2. CRHR1 is a genome-wide significant locus for bone-mineral-density [46] and a suggestive locus for infant head circumference [47], and it is also involved in stress, reproduction, immune response, and obesity [48, 49]. NOS3 is involved in several sex-specific molecular processes, including cardiac and endothelial function [50, 51]. FARS2 is a suggestive locus (p<10−7) for severe early-onset obesity [52].

The investigation of WHR data indicated putative signals of polygenetic adaptation in the systems genetics of this trait. In particular, the female-specific findings suggest interplay between lipid metabolism and immune systems in the enrichment analysis. The most immediately relevant locus involved in this gene set with evidence polygenic adaptation is PPARG. This gene was genome-wide significant in a GWAS of WHR in women [25]. Furthermore, population genetics studies indicated that it is a candidate in relation to metabolic adaptation [53]. Finally, investigations using different approaches uncovered several functions of this gene and its involvement in the pathogenesis of obesity, diabetes, atherosclerosis, and cancer [54]. The enrichment analysis based on the PPARG gene and the other genes involved in polygenic adaptation related to female WHR (i.e., IKBKAP, APOL1, HLA-B, and MED9) indicated an overlapping between immune systems and lipid transportation. Both these molecular processes are widely recognized as mechanisms under selective pressures during human evolution [13, 55], and molecular investigations have also highlighted the strong interplay between them [56, 57]. Furthermore, the sexes showed significant differences in both these mechanisms, mainly attributable to the female reproductive function [58, 59]. All these results are consistent with the possibility of polygenic adaption observed in the female WHR gene set.

In conclusion, our study suggests an effective approach to the investigation of polygenic adaption in the systems genetic of complex traits based on combining GWAS data with information regarding the systems genetics of complex traits. Regarding the results obtained about anthropometric traits, our findings lead to novel insights that indicate polygenic adaptation in response to selective pressures related to locomotory behavior, infection resistance, and lipid transport. Further, in sex-stratified analysis of WC and WHR, we observed most of the significant findings in females. These data indicate that the effect of past adaptation processes (e.g., locomotory behaviors and infection resistance) on human genome variation are affecting the predisposition to anthropometric traits. However, these findings are specific to the genetic structure of CEU populations that is the most used reference panel for European populations. Additional polygenic adaptation processes are expected for other European and non-European populations.

Besides its potential contribution to population genetics, our study also suggested the understanding of epistatic interactions related to polygenic adaptation may also improve our knowledge about systems genetics, highlighting novel putative interactive processes to investigate. Finally, here we focused on anthropometric traits, obtaining interesting results about the potential role of human adaptation in shaping the genetic predisposition to these traits. However, numerous GWAS complex traits datasets are available to be investigated. The exploration of these data may provide further information useful to understand both the genetic mechanisms of human adaptation, and their effects on the genetic predisposition to complex traits.

Supporting Information

S1 Table. GWAS data of anthropometric traits made available by GIANT consortium that were used in the present study.

https://doi.org/10.1371/journal.pone.0160654.s001

(DOCX)

S2 Table. gene networks associated with GIANT phenotypic traits that were constructed on the basis of top-10 PPI modules constructed with variants with |iHS| > 1.5.

https://doi.org/10.1371/journal.pone.0160654.s002

(DOCX)

S3 Table. Genes and their correspondence p values present in gene network associated with BMI-related phenotypes.

NI: not included in top-10 PPI modules.

https://doi.org/10.1371/journal.pone.0160654.s003

(DOCX)

S4 Table. Significant long-distance genotypic LDs observed in BMI-associated gene networks.

https://doi.org/10.1371/journal.pone.0160654.s004

(DOCX)

S5 Table. Genes and their correspondence p values present in gene network associated with height-related phenotypes.

NI: not included in top-10 PPI modules.

https://doi.org/10.1371/journal.pone.0160654.s005

(DOCX)

S6 Table. Significant long-distance genotypic LDs observed in Height-associated gene networks.

https://doi.org/10.1371/journal.pone.0160654.s006

(DOCX)

S7 Table. Genes and their correspondence p values present in gene network associated with WC-related phenotypes.

NI: not included in top-10 PPI modules.

https://doi.org/10.1371/journal.pone.0160654.s007

(DOCX)

S8 Table. Significant long-distance genotypic LDs observed in WC-associated gene networks.

https://doi.org/10.1371/journal.pone.0160654.s008

(DOCX)

S9 Table. Genes and their correspondence p values present in gene network associated with WHR-related phenotypes.

NI: not included in top-10 PPI modules.

https://doi.org/10.1371/journal.pone.0160654.s009

(DOCX)

S10 Table. Significant long-distance genotypic LDs observed in WHR-associated gene networks.

https://doi.org/10.1371/journal.pone.0160654.s010

(DOCX)

Acknowledgments

The authors are grateful to the research groups of the GIANT consortium and the HapMap Project for their publicly available data.

Author Contributions

  1. Conceived and designed the experiments: RP.
  2. Performed the experiments: RP.
  3. Analyzed the data: RP HZ.
  4. Contributed reagents/materials/analysis tools: RP JG.
  5. Wrote the paper: RP.
  6. Data interpretation: RP JG. Manuscript revision for important intellectual content: BZY HZ JG. Final approval: RP BZY HZ JG.

References

  1. 1. Lachance J, Tishkoff SA. Population Genomics of Human Adaptation. Annu Rev Ecol Evol Syst. 2013;44:123–43. pmid:25383060
  2. 2. Jeong C, Di Rienzo A. Adaptations to local environments in modern human populations. Curr Opin Genet Dev. 2014;29C:1–8.
  3. 3. Colonna V, Ayub Q, Chen Y, Pagani L, Luisi P, Pybus M, et al. Human genomic regions with exceptionally high levels of population differentiation identified from 911 whole-genome sequences. Genome Biol. 2014;15:R88. pmid:24980144
  4. 4. Fumagalli M, Sironi M. Human genome variability, natural selection and infectious diseases. Curr Opin Immunol. 2014;30:9–16. pmid:24880709
  5. 5. Jablonski NG. The evolution of human skin colouration and its relevance to health in the modern world. J R Coll Physicians Edinb. 2012;42:58–63. pmid:22441067
  6. 6. Mortensen HM, Froment A, Lema G, Bodo JM, Ibrahim M, Nyambo TB, et al. Characterization of genetic variation and natural selection at the arylamine N-acetyltransferase genes in global human populations. Pharmacogenomics. 2011;12:1545–58. pmid:21995608
  7. 7. Wuren T, Simonson TS, Qin G, Xing J, Huff CD, Witherspoon DJ, et al. Shared and unique signals of high-altitude adaptation in geographically distinct Tibetan populations. PLoS One. 2014;9:e88252. pmid:24642866
  8. 8. Pritchard JK, Di Rienzo A. Adaptation—not by sweeps alone. Nat Rev Genet. 2010;11:665–7. pmid:20838407
  9. 9. Pritchard JK, Pickrell JK, Coop G. The genetics of human adaptation: hard sweeps, soft sweeps, and polygenic adaptation. Curr Biol. 2010;20:R208–15. pmid:20178769
  10. 10. Hou L, Zhao H. A review of post-GWAS prioritization approaches. Front Genet. 2013;4:280. pmid:24367376
  11. 11. Polimanti R, Yang C, Zhao H, Gelernter J. Dissecting ancestry genomic background in substance dependence genome-wide association studies. Pharmacogenomics. 2015;16:1487–98. pmid:26267224
  12. 12. Turchin MC, Chiang CW, Palmer CD, Sankararaman S, Reich D, Genetic Investigation of ATC, et al. Evidence of widespread selection on standing variation in Europe at height-associated SNPs. Nat Genet. 2012;44:1015–9. pmid:22902787
  13. 13. Daub JT, Hofer T, Cutivet E, Dupanloup I, Quintana-Murci L, Robinson-Rechavi M, et al. Evidence for polygenic adaptation to pathogens in the human genome. Mol Biol Evol. 2013;30:1544–58. pmid:23625889
  14. 14. Berg JJ, Coop G. A population genetic signal of polygenic adaptation. PLoS Genet. 2014;10:e1004412. pmid:25102153
  15. 15. Wells JC. Ecogeographical associations between climate and human body composition: analyses based on anthropometry and skinfolds. Am J Phys Anthropol. 2012;147:169–86. pmid:22212891
  16. 16. Courtiol A, Rickard IJ, Lummaa V, Prentice AM, Fulford AJ, Stearns SC. The demographic transition influences variance in fitness and selection on height and BMI in rural Gambia. Curr Biol. 2013;23:884–9. pmid:23623548
  17. 17. Goldstein MS, Kobyliansky E. Anthropometric traits, balanced selection and fertility. Hum Biol. 1984;56:35–46. pmid:6745902
  18. 18. Chan Y, Lim ET, Sandholm N, Wang SR, McKnight AJ, Ripke S, et al. An excess of risk-increasing low-frequency variants can be a signal of polygenic inheritance in complex diseases. Am J Hum Genet. 2014;94:437–52. pmid:24607388
  19. 19. Ferrer-Admetlla A, Liang M, Korneliussen T, Nielsen R. On detecting incomplete soft or hard selective sweeps using haplotype structure. Mol Biol Evol. 2014;31:1275–91. pmid:24554778
  20. 20. Speliotes EK, Willer CJ, Berndt SI, Monda KL, Thorleifsson G, Jackson AU, et al. Association analyses of 249,796 individuals reveal 18 new loci associated with body mass index. Nat Genet. 2010;42:937–48. pmid:20935630
  21. 21. Yang J, Loos RJ, Powell JE, Medland SE, Speliotes EK, Chasman DI, et al. FTO genotype is associated with phenotypic variability of body mass index. Nature. 2012;490:267–72. pmid:22982992
  22. 22. Berndt SI, Gustafsson S, Magi R, Ganna A, Wheeler E, Feitosa MF, et al. Genome-wide meta-analysis identifies 11 new loci for anthropometric traits and provides insights into genetic architecture. Nat Genet. 2013;45:501–12. pmid:23563607
  23. 23. Lango Allen H, Estrada K, Lettre G, Berndt SI, Weedon MN, Rivadeneira F, et al. Hundreds of variants clustered in genomic loci and biological pathways affect human height. Nature. 2010;467:832–8. pmid:20881960
  24. 24. Heid IM, Jackson AU, Randall JC, Winkler TW, Qi L, Steinthorsdottir V, et al. Meta-analysis identifies 13 new loci associated with waist-hip ratio and reveals sexual dimorphism in the genetic basis of fat distribution. Nat Genet. 2010;42:949–60. pmid:20935629
  25. 25. Randall JC, Winkler TW, Kutalik Z, Berndt SI, Jackson AU, Monda KL, et al. Sex-stratified genome-wide association studies including 270,000 individuals show sexual dimorphism in genetic loci for anthropometric traits. PLoS Genet. 2013;9:e1003500. pmid:23754948
  26. 26. Voight BF, Kudaravalli S, Wen X, Pritchard JK. A map of recent positive selection in the human genome. PLoS Biol. 2006;4:e72. pmid:16494531
  27. 27. Ding K, Kullo IJ. Geographic differences in allele frequencies of susceptibility SNPs for cardiovascular disease. BMC Med Genet. 2011;12:55. pmid:21507254
  28. 28. Southam L, Soranzo N, Montgomery SB, Frayling TM, McCarthy MI, Barroso I, et al. Is the thrifty genotype hypothesis supported by evidence based on confirmed type 2 diabetes- and obesity-susceptibility variants? Diabetologia. 2009;52:1846–51. pmid:19526209
  29. 29. Polimanti R, Piacentini S, Manfellotto D, Fuciarelli M. Human genetic variation of CYP450 superfamily: analysis of functional diversity in worldwide populations. Pharmacogenomics. 2012;13:1951–60. pmid:23215887
  30. 30. Choudhury A, Hazelhurst S, Meintjes A, Achinike-Oduaran O, Aron S, Gamieldien J, et al. Population-specific common SNPs reflect demographic histories and highlight regions of genomic plasticity with functional relevance. BMC Genomics. 2014;15:437. pmid:24906912
  31. 31. Liu JZ, McRae AF, Nyholt DR, Medland SE, Wray NR, Brown KM, et al. A versatile gene-based test for genome-wide association studies. Am J Hum Genet. 2010;87:139–45. pmid:20598278
  32. 32. Jia P, Zheng S, Long J, Zheng W, Zhao Z. dmGWAS: dense module searching for genome-wide association studies in protein-protein interaction networks. Bioinformatics. 2011;27:95–102. pmid:21045073
  33. 33. Cowley MJ, Pinese M, Kassahn KS, Waddell N, Pearson JV, Grimmond SM, et al. PINA v2.0: mining interactome modules. Nucleic Acids Res. 2012;40:D862–5. pmid:22067443
  34. 34. Rousset F. genepop'007: a complete re-implementation of the genepop software for Windows and Linux. Mol Ecol Resour. 2008;8:103–6. pmid:21585727
  35. 35. 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. pmid:17701901
  36. 36. International HapMap Consortium, Altshuler DM, Gibbs RA, Peltonen L, Altshuler DM, Gibbs RA, et al. Integrating common and rare genetic variation in diverse human populations. Nature. 2010;467:52–8. pmid:20811451
  37. 37. Huang da W, Sherman BT, Lempicki RA. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 2009;37:1–13. pmid:19033363
  38. 38. Tsalik EL, Hobert O. Functional mapping of neurons that control locomotory behavior in Caenorhabditis elegans. J Neurobiol. 2003;56:178–97. pmid:12838583
  39. 39. Sakkou M, Wiedmer P, Anlag K, Hamm A, Seuntjens E, Ettwiller L, et al. A role for brain-specific homeobox factor Bsx in the control of hyperphagia and locomotory behavior. Cell Metab. 2007;5:450–63. pmid:17550780
  40. 40. Zahra K, Khan MA, Iqbal F. Oral supplementation of Ocimum basilicum has the potential to improves the locomotory, exploratory, anxiolytic behavior and learning in adult male albino mice. Neurol Sci. 2014.
  41. 41. Polimanti R, Zhang H, Smith AH, Zhao H, Farrer LA, Kranzler HR, et al. Genome-wide association study of body mass index in subjects with alcohol dependence. Addict Biol. 2015.
  42. 42. Wood AR, Esko T, Yang J, Vedantam S, Pers TH, Gustafsson S, et al. Defining the role of common variation in the genomic and biological architecture of adult human height. Nat Genet. 2014;46:1173–86. pmid:25282103
  43. 43. Limou S, Le Clerc S, Coulonges C, Carpentier W, Dina C, Delaneau O, et al. Genomewide association study of an AIDS-nonprogression cohort emphasizes the role played by HLA genes (ANRS Genomewide Association Study 02). J Infect Dis. 2009;199:419–26. pmid:19115949
  44. 44. Piwoz E, Sundberg S, Rooke J. Promoting healthy growth: what are the priorities for research and action? Adv Nutr. 2012;3:234–41. pmid:22516734
  45. 45. Joel DR, Mabikwa V, Makhanda J, Tolle MA, Anabwani GM, Ahmed SF. The prevalence and determinants of short stature in HIV-infected children. J Int Assoc Provid AIDS Care. 2014;13:529–33. pmid:24752886
  46. 46. Rivadeneira F, Styrkarsdottir U, Estrada K, Halldorsson BV, Hsu YH, Richards JB, et al. Twenty bone-mineral-density loci identified by large-scale meta-analysis of genome-wide association studies. Nat Genet. 2009;41:1199–206. pmid:19801982
  47. 47. Taal HR, St Pourcain B, Thiering E, Das S, Mook-Kanamori DO, Warrington NM, et al. Common variants at 12q15 and 12q24 are associated with infant head circumference. Nat Genet. 2012;44:532–8. pmid:22504419
  48. 48. Gragnoli C. Hypothesis of the neuroendocrine cortisol pathway gene role in the comorbidity of depression, type 2 diabetes, and metabolic syndrome. Appl Clin Genet. 2014;7:43–53. pmid:24817815
  49. 49. Zoumakis E, Kalantaridou SN, Makrigiannakis A. CRH-like peptides in human reproduction. Curr Med Chem. 2009;16:4230–5. pmid:19754421
  50. 50. Loria AS, Brinson KN, Fox BM, Sullivan JC. Sex-specific alterations in NOS regulation of vascular function in aorta and mesenteric arteries from spontaneously hypertensive rats compared to Wistar Kyoto rats. Physiol Rep. 2014;2.
  51. 51. Zuloaga KL, Davis CM, Zhang W, Alkayed NJ. Role of aromatase in sex-specific cerebrovascular endothelial function in mice. Am J Physiol Heart Circ Physiol. 2014;306:H929–37. pmid:24508640
  52. 52. Wheeler E, Huang N, Bochukova EG, Keogh JM, Lindsay S, Garg S, et al. Genome-wide SNP and CNV analysis identifies common and low-frequency variants associated with severe early-onset obesity. Nat Genet. 2013;45:513–7. pmid:23563609
  53. 53. Xing J, Wuren T, Simonson TS, Watkins WS, Witherspoon DJ, Wu W, et al. Genomic analysis of natural selection and phenotypic variation in high-altitude mongolians. PLoS Genet. 2013;9:e1003634. pmid:23874230
  54. 54. Janani C, Ranjitha Kumari BD. PPAR gamma gene—A review. Diabetes Metab Syndr. 2014.
  55. 55. Babin PJ, Gibbons GF. The evolution of plasma cholesterol: direct utility or a "spandrel" of hepatic lipid metabolism? Prog Lipid Res. 2009;48:73–91. pmid:19049814
  56. 56. Festi D, Schiumerini R, Eusebi LH, Marasco G, Taddia M, Colecchia A. Gut microbiota and metabolic syndrome. World J Gastroenterol. 2014;20:16079–94. pmid:25473159
  57. 57. Legein B, Temmerman L, Biessen EA, Lutgens E. Inflammation and immune system interactions in atherosclerosis. Cell Mol Life Sci. 2013;70:3847–69. pmid:23430000
  58. 58. Henderson GC. Sexual dimorphism in the effects of exercise on metabolism of lipids to support resting metabolism. Front Endocrinol (Lausanne). 2014;5:162.
  59. 59. Houser BL. Decidual macrophages and their roles at the maternal-fetal interface. Yale J Biol Med. 2012;85:105–18. pmid:22461749