Genome-wide association scans (GWAS) have been extraordinarily successful in identifying common DNA sequence variants (single nucleotide polymorphisms, SNPs) associated with diseases, risk factors and complex phenotypes [1–6]. Increasingly large GWAS have expanded our knowledge of specific genes and pathways influencing lipid levels. Technological advances in DNA sequencing have made the interrogation of protein-coding regions of the genome (the exome) affordable. Although exome sequencing has been particularly useful in identifying genes for Mendelian forms of disease [7,8], for more complex human phenotypes, including lipids, it has been less successful. The current status of genetic investigation in lipid traits is the subject of this review.
GENOME-WIDE ASSOCIATION SCANS
GWAS for blood lipid levels have identified hundreds of novel genomic loci that are also associated with cardiovascular and other diseases. The most recent GWAS for blood lipids was a study of nearly 187 000 individuals by the Global Lipids Genetics Consortium (GLGC) [9▪▪]. The GLGC identified 157 independent loci significantly associated with at least one of the four lipid traits analysed, HDL cholesterol (HDL-C), LDL cholesterol (LDL-C), total cholesterol and triglyceride levels (Fig. 1). Multiple computational approaches, including assessing variants for association with gene expression, identification of coding variants in linkage disequilibrium, literature text mining, protein–protein interaction and network analyses, were applied to identify the most likely candidate functional genes and variants. Confirmation of biologically functional genes and mechanisms related to lipid traits will require laboratory-based functional experiments in cellular or model systems.
New loci discovered by GWAS typically have small effect sizes (representing changes of a few mg/dl by genotype) with a correspondingly small proportion of trait variance explained by individual variants (typically <0.1%).Nonetheless, the genes at these loci may provide novel biological insight and drug targets, best illustrated by the HMG Co-A Reductase (HMGCR) gene [10–12]. Statins, which target HMG Co-A reductase, have a dramatic impact on blood lipid levels (30–50% reduction of LDL-C) with similar decreases in risk of cardiac events. Using a single variant in HMGCR as a proxy for HMGCR-inhibition by statins, a meta-analysis demonstrated a mean 1.08 mg/dl lower LDL-C level with each additional minor allele (rs17238484-G); yet, statins have major effects on LDL-C and disease risk [13▪▪]. Thus, variants associated with a trait may have a small effect, yet the gene/pathway could be a drug target with a dramatic impact on the trait and associated disease risk.
Genome-wide association scans: lipid variants identifying therapeutic targets
Many SNPs associated with human phenotypes are located in noncoding regions of the genome , with limited understanding of how these variants affect phenotypes and disease risk. The ENCODE (ENCyclopedia Of DNA Elements) project was established to identify functional elements in the human genome and has generated a diverse and large-scale array of various predictors of functionality (e.g. DNase hypersensitivity, histone methylation marks, transcription factor binding sites) in a wide variety of cell types representing many human tissues . As in any prediction, experimental validation in the laboratory is necessary, as shown with a GWAS SNP in the regulatory region of the SORT1 gene . The translation from gene discovery to therapeutic targets is illustrated by statins (HMGCR) and ezetimibe (NPC1L1). Interest in pharmacologically based PCSK9 inhibitors emerged from the discovery of loss-of-function variants in the PCSK9 gene that result in substantially lower LDL-C levels [18,19▪]. As with any development of drugs related to biological targets, the timeframe required for development of novel therapeutics may be long and the federal approval path uncertain.
Using lipid genome-wide association scans: single nucleotide polymorphisms to assess causality of lipids on coronary artery disease
Identification of genetic variants associated with lipids permits testing for a causal role of lipid trait variation on subsequent disease risk. Mendelian randomization , or more correctly called multivariate instrumental variable test, suggests that HDL-C is not a causal risk factor for heart disease , consistent with the failure of HDL-C raising drugs to lower heart disease events. Genetic variants that have a large impact on LDL-C level should have a large impact on risk of coronary artery disease (CAD) through the CAD–LDL-C association. An advantage of this analytic approach is that one can adjust for the impact on other traits. Thus, the relationship between the variants’ impacts on LDL-C, and their impacts on CAD, remain strongly significant, even after adjustment for the variants’ impacts on HDL-C or triglyceride levels. For HDL-C, the correlation between the impact of the variants on HDL-C and on CAD is weakly significant; after adjusting for the impacts on triglycerides, there is little correlation remaining between the impacts on HDL-C and on CAD. For variants that affect triglyceride levels, there is a strong correlation with impact on CAD that is not attenuated by inclusion of the impact on HDL-C, suggesting that triglycerides, and not HDL-C, may be causal for CAD.
EXOME SEQUENCING FOR LIPID TRAITS
Prior to GWAS, linkage and targeted sequencing studies were used to identify a small number of genes with highly penetrant coding (putatively functional) mutations that segregated with extreme lipid measures in some families. Although GWAS and linkage have been successful for identifying lipid-associated loci, associated variants explain only a small amount of the heritability for lipid traits. This suggests that there could be additional genes and variants with large effects on lipid levels that have yet to be identified. Exome sequencing discovers and tests for association of rare variants not well represented on GWAS panels. Exome sequencing studies have been performed both on participants selected from families with extreme lipid levels and on participants from population-based studies.
Exome sequencing for familial dyslipidemias
Three recent studies, highlighted in Table 1[22–25,26▪▪], identified novel causal genes in single families with familial combined hypolipidemia (FCH) [22,27], autosomal recessive hypercholesterolemia (ARH)  and hypertriglyceridemia . FCH is characterized by very low plasma levels of LDL-C, HDL-C and triglycerides and typically is caused by mutations in the APOB gene. An FCH sibling pair from a large family that excluded linkage to APOB was used to identify two shared, previously unidentified, variants in ANGPTL3; in this family, the HDL-C levels are inherited in a pattern consistent with a single gene recessive model. Sanger sequencing ANGPTL3 in all 38 members of the family confirmed that the two nonsense mutations segregated independently, suggesting that the two siblings were compound heterozygotes. Family members with neither mutation had the highest LDL-C, those with one mutation had intermediate LDL-C and four members with compound heterozygotes had the lowest LDL-C.
A similar approach was applied to a family with a proband and two affected siblings (one of which was a monozygotic twin to the proband) with ARH , a disorder characterized by extremely high total cholesterol and LDL-C (Table 1). All affected siblings had LDL-C greater than the 99th percentile, whereas the father and mother had ‘normal’ LDL-C, suggesting an autosomal recessive mode of inheritance. Exome sequencing was performed on the proband, his nontwin brother and ‘normal’ father. Only two missense variants were identified, one in exon 8 in LIPA, previously implicated in cholesterol ester storage disease. It was determined that the siblings had an unusual presentation of cholesterol ester storage disease that was previously misdiagnosed, highlighting the utility of exome sequencing to inform clinical diagnosis.
Exome sequencing was performed on 16 family members from a 221 person, five-generation family ascertained for FCHL to identify novel genes/variants influencing triglycerides . In these family members, the hypertriglyceridemia is not severe, but mild to moderate. From exome analysis, two genes were interrogated by testing variants in the complete family. Five shared, highly conserved, private missense variants in SLC25A40 and PLD2 jointly explained nearly 50% of the genetic variance of triglyceride levels (Table 1). One SLC25A40 variant (c.374A>G) was predicted to cause a disruptive p.Tyr125Cys substitution near the second helical transmembrane region of the SLC25A40 inner mitochondrial membrane transport protein.
Exome sequencing for genes contributing to variation in lipids
A challenge of studying unrelated individuals, rather than multiple members of large Mendelian families, is that most unrelated individuals will likely carry different causal coding variants that give rise to a similar phenotype. In the absence of a strong founder effect or a variant under strong selective pressure, the power to detect association for most individual causal variants will be small. Analytic methods have been developed to assess the association between the aggregate burden of rare variant alleles on traits such as LDL-C [28,29]. This approach can increase statistical power by increasing the number of individuals carrying the risk-altering exposure (i.e. the number of individuals carrying any putatively important mutation) and by decreasing the multiple test burden applied to the hypothesis test results, as the number of tests is now defined by the number of genes rather than the number of variants.
The NHLBI Exome Sequencing Project conducted the largest exome sequencing study of LDL-C to date [26▪▪]. In a multistage design, 3307 European American (47%) and African-American (53%) participants from seven cohorts were examined, including 855 individuals selected for extreme high or low LDL-C. TOMM40 SNP rs1160983 was significantly associated with LDL-C. No APOE SNPs, including rs7412 and rs429358 that tag the classical APOE isoforms, passed quality control metrics. The TOMM40 SNP rs1160983 is highly correlated with APOE rs7412 (p.Arg176Cys) with D’ = 0.63. The minor allele of rs1160983 tags the APOE-E2 isoform (nearly two-thirds of the time) and the APOE-E3 isoform (nearly one-third of the time) but not the APOE-E4 isoform. These data suggested that the TOMM40 effect could be accounted in large part by linkage disequilibrium with APOE. Meta-analysis of the two-stage sample identified four genes (LDLR, APOB, PCSK9 and PNPLA5) significantly associated with LDL-C, representing one novel gene (PNPLA5) and three known genes with novel variants (Table 1). The novel finding for a burden of PNPLA5 nonsynonymous and splice site variants with LDL-C was replicated in 2084 individuals from the GoT2D (Genetics of T2D) consortium.
A burden of rare mutations in APOC3 was associated with lower plasma triglyceride levels in 3734 participants of European American and African-American ancestry in the NHLBI Exome Sequencing Project [30▪▪]. Among four critical mutations, three were loss-of-function and the fourth was a missense mutation. Triglyceride levels in the carriers were 39% lower than in noncarriers and circulating levels of APOC3 in carriers were 46% lower than in noncarriers. The risk of coronary heart disease among 498 carriers with any rare APOC3 mutation was 40% lower than the risk among noncarriers (N = 110 472), suggesting a biological pathway through APOC3 that not only affects triglyceride levels but also coronary disease risk. These results have been observed in two population-based studies in Denmark (N = 75 725) with heterozygosity for any of three loss-of-function mutations in APOC3 being associated with low levels of triglyerides as well as reduced risk of ischemic vascular disease . The rare APOC3 R19X variant occurs in many populations, yet it consistently lowers triglyceride levels and reduces cardiovascular disease risk. This suggests positive selective pressure to maintain the R19X variant across populations, even if rare.
A smaller exome sequencing study was conducted on 125 unrelated individuals with definite familial hypercholesterolemia (DFH) and 1926 controls from the UK10K project . DFH cases were examined for mutations in genes known to have variants associated with DFH (LDLR, APOB, PCSK9, LDLRAP1), identifying 23 individuals with LDLR mutations and two individuals with APOB mutations who were missed by standard genetic screening protocols. A genetic risk score based on 12 established LDL-C increasing variants in or near 11 genes [32▪] identified 29 participants with ‘polygenic’ familial hypercholesterolemia (DFH cases). DFH cases without mutations in the four known DFH genes had significantly higher mean genetic risk scores than controls or DFH cases with mutations in known DFH genes. Exome sequencing was performed, excluding the 25 DFH cases with mutations in known DFH genes and 29 cases with high gene risk scores. In 71 DFH cases and 1926 controls, three rare loss-of-function variants were found in the cholesterol 25-hydroxylase (CH25H) gene that catalyzes the formation of oxysterol – 25-hydroxycholesterol and is in the same pathway of cholesterol metabolism.
Genotyping using arrays containing rare coding-region variants
A low-cost genotyping array, generically termed the ‘ExomeChip’, was designed to capture coding and splice site variation for infrequent variants from 12 000 participants with exome or whole genome sequence data. The ExomeChip, however, cannot capture the numerous rare mutations. A meta-analysis of ExomeChip data from 56 538 individuals (42 208 European American and 14 330 African-American) was performed for association with LDL-C, HDL-C and triglyceride levels [33▪▪]. Single variant tests identified novel, statistically significant associations: ANGPTL8 with HDL-C level in European American (rs145464906; c.361C>T; p.Gln121), PAFAH1B2with HDL-C and triglyceride levels in European American (rs186808413; c.482C>T; p.Ser161Leu), COL18A1 with triglyceride levels in African-American (rs114139997; c.331G>A; p.Gly111Arg) and PCSK7 with HDL-C and triglyceride levels in African-American (rs142953140; c.1511G>A; p.Arg504His). New associations were not observed for LDL-C with either single variant or gene-burden tests.
A subsequent analysis of LDL-C, HDL-C, triglyceride and total cholesterol levels was conducted in 5643 Norwegians, followed by genotyping the most associated 18 variants in a replication sample of 4666 Norwegians . Ten variants attained genome-wide significance in the combined sample, with nine located in or near well established lipid-related genes and the tenth (TM6SF2; p.Glu167Lys) in a previously identified NCAN/CILP2/PBX4/19p13 GWAS locus (rs10401969) associated with total cholesterol . In 20 597 CAD cases and 61 046 controls, the TM6SF2 rs10401969 variant was associated with a decreased risk of disease. A series of functional studies demonstrated that endogenous Tm6sf2 was highly expressed in the liver of C75BL/6J mice, and that transient TM6SF2 overexpression and knockdown of Tm6sf2 altered serum lipid profiles in a direction consistent with the observed associations in humans, and replicated in the Dallas Heart Study .
Other studies using the ExomeChip have been conducted in a variety of populations. A population isolate of 1267 Greek individuals identified evidence for association for the nonsynonymous APOC3 R19X variant with increased HDL-C and decreased triglycerides, explaining a substantial component of variation when compared with GWAS associations [37▪]. The APOC3 R19X variant was thought to be a founder variant in the Amish , yet it may be more widespread due to its functional impact. These results suggest that interrogation of isolated populations may be a useful approach to identify variants that may be common in the isolates and associated with complex human traits in the general population.
GWAS, coupled with meta-analytic approaches, have been remarkably successful in identifying common DNA sequence variants, many located in regulatory regions of genes, which are robustly associated with variation in lipid phenotypes (HDL-C, LDL-C, total cholesterol, triglycerides). The GLGC identified 157 independent loci robustly associated with at least one of these four lipid traits. Exome sequencing has been employed to identify low-frequency variants in protein-coding regions of genes with highly penetrant effects on lipid traits. These results demonstrate that the novel low-frequency variants often have a greater impact than the common variants identified by GWAS (Fig. 2). However, even with the use of sequencing and genotyping arrays targeting rare/infrequent variants, there have been relatively few new genes convincingly identified that impact lipid levels with the sample sizes analysed to date.
Currently, genetic studies have produced a large number of gene targets for lipids, the majority with a small effect and many in regulatory regions of genes. These genes will require further study of their variants, impact on function and role on regulation. The next stage, logically, for comprehensive biological analysis of lipids is whole genome sequencing. This process will complete the assessment of the remaining 99% of the human genome for association with variation in lipid levels. Although the interpretation of these currently unknown variants may be daunting, the integration of these data with data being generated by transcriptomic and epigenomic projects may facilitate our rapid increase in knowledge of lipid biology and development of therapeutic targets to provide improved lipid homeostasis.
This work was supported by grants from the National Institutes of Health, NHLBI Grand Opportunity Exome Sequencing Project (ESP) RC2 HL-103010 (S.S.R. and L.A.L.), HV-13-09 (L.A.L), HL071862 (L.A.L), HL094535 (C.J.W.) and HL109946 (C.J.W.).
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
REFERENCES AND RECOMMENDED READING
Papers of particular interest, published within the annual period of review, have been highlighted as:
- ▪ of special interest
- ▪▪ of outstanding interest
1. Psaty BM, O’Donnell CJ, Gudnason V, et al. Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE): design of prospective meta-analyses of genome-wide association studies. Circ Cardiovasc Genet 2009; 2:73–80.
2. Teslovich TM, Musununu K, Smith AV, et al. Biological, clinical, and population relevance of 95 loci mapped for serum lipid concentrations. Nature 2010; 466:707–713.
3. Lango Allen H, Estrada K, Lettre G, et al. Hundreds of variants influence human height and cluster within genomic loci and biological pathways. Nature 2010; 467:832–838.
4. Speliotes EK, Willer CJ, Berndt SI, et al. Association analyses of 249 796 individuals reveals 18 new loci associated with body mass index. Nat Genet 2010; 42:937–948.
5. O’Donnell CJ, Kavousi M, Smith AV, et al. Genome-wide association study for coronary artery calcium with follow-up in myocardial infarction. Circulation 2011; 124:2855–2864.
6. Schunkert H, Konig IR, Kathiresan S, et al. Large-scale association analysis identified 13 new susceptiblity loci for coronary artery disease. Nat Genet 2011; 43:333–338.
7. Ng SB, Nickerson DA, Bamshad MJ, Shendure J. Massively parallel sequencing and rare disease. Hum Mol Genet 2010; 19:R119–R124.
8. Bamshad MJ, Ng SB, Bigham AW, et al. Exome sequencing
as a tool for Mendelian disease gene discovery. Nat Rev Genet 2011; 12:745–755.
9▪▪. Willer CJ, Schmidt EM, Sengupta S, et al. Global Lipids Genetics Consortium. Discovery and refinement of loci associated with lipid levels. Nat Genet 2013; 45:1274–1283.
A meta-analysis of 188 577 individuals from multiple cohorts to discover new loci and to reduce and refine the location of existing loci associated with LDL-C, HDL-C, triglyceride and total cholesterol levels in multiple ethnicities and using multiple analytic and bioinformatics tools. An important exhibition of using genomic and functional annotations to define potential biological mechanisms and a useful overview of the tools of exome sequencing and the approach to the approach of the discovery of novel genes that lead to single gene disorders.
10. Polisecki E, Muallem H, Maeda N, et al. Genetic variation at the LDL receptor and HMG-CoA reductase gene loci, lipid levels, statin response, and cardiovascular disease incidence in PROSPER. Atherosclerosis 2008; 200:109–114.
11. Donnelly LA, Doney AS, Dannfald J, et al. A paucimorphic variant in the HMG-CoA reductase gene is associated with lipid-lowering response to statin treatment in diabetes: a GoDARTS study. Pharmacogenet Genomics 2008; 18:1021–1026.
12. Mangravite LM, Medina MW, Cui J, et al. Combined influence of LDLR and HMGCR sequence variation on lipid-lowering response to simvastatin. Arterioscler Thromb Vasc Biol 2010; 30:1485–1492.
13▪▪. Swerdlow DI, Preiss D, Kuchenbaecker KB, et al. HMG-coenzyme A reductase inhibition, type 2 diabetes, and bodyweight: evidence from genetic analysis and randomised trials. Lancet 2015; 385:24–30.
A meta-analytic approach to evaluate the role of the primary HMGCR SNP (rs17238484) and its proxy (rs12916) in 223 463 participants in 43 studies for the role of the variant on lipid, obesity and diabetes phenotypes with respect to effect of statin use. The analyses confirm the role of the variant on specific phenotypes but suggest that the variant and HMGCR inhibition is partially responsible for the increased risk of type 2 diabetes while decreasing lipids.
14. Maurano MT, Humbert R, Rynes E, et al. Systematic localization of common disease-associated variation in regulatory DNA. Science 2012; 337:1190–1195.
15. Kellis M, Wold B, Snyder MP, et al. Defining functional DNA elements in the human genome. Proc Natl Acad Sci U S A 2014; 111:6131–6138.
16. Musunuru K, Strong A, Frank-Kamenetsky M, et al. From noncoding variant to phenotype via SORT1 at the 1p13 cholesterol locus. Nature 2010; 466:714–719.
17. Wang J, Williams CM, Hegele RA. Compound heterozygosity for two nonsynonymous polymorphisms in NPC1L1 in a nonresponder to ezetimibe. Clin Genet 2005; 67:175–177.
18. Abifadel M, Varret M, Rabes JP, et al. Mutations in PCSK9 cause autosomal dominant hypercholesterolemia. Nat Genet 2003; 34:154–156.
19▪. Cohen JC, Hobbs HH. Simple genetics for a complex disease. Science 2013; 340:689–690.
A useful overview of the PCSK9 story, from gene discovery of its association with LDL-C, to the steps in translating that discovery to therapeutic target for human use.
20. Trompet S, Jekema JW, Katan MB, et al. Apolipoprotein e genotype, plasma cholesterol, and cancer: a Mendelian randomization study. Am J Epidemiol 2009; 170:1415–1421.
21. Voight BF, Peloso GM, Orho-Melander M, et al. Plasma HDL cholesterol and risk for myocardial infarction: a Mendelian randomization study. Lancet 2012; 380:572–580.
22. Musunuru K, Pirruccello JP, Do R, et al. Exome sequencing
, ANGPTL3 mutations, and familial combined hypolipidemia. N Engl J Med 2010; 363:2220–2227.
23. Stitziel NO, Fouchier SW, Sjouke B, et al. Exome sequencing
and directed clinical phenotyping diagnose cholesterol ester storage disease presenting as autosomal recessive hypercholesterolemia. Arterioscler Thromb Vasc Biol 2013; 33:2909–2914.
24. Futema M, Plagnol V, Li K, et al. Whole exome sequencing
of familial hypercholesterolaemia patients negative for LDLR/APOB/PCSK9 mutations. J Med Genet 2014; 51:537–544.
25. Rosenthal EA, Ranchalis J, Crosslin DR, et al. Joint linkage and association analysis with exome sequence data implicates SLC25A40 in hypertriglyceridemia. Am J Hum Genet 2013; 93:1035–1045.
26▪▪. Lange LA, Hu Y, Zhang H, et al. Whole-exome sequencing
identifies rare and low-frequency coding variants associated with LDL cholesterol. Am J Hum Genet 2014; 94:233–245.
The first major study on exome sequencing for LDL-C, providing detailed approaches to the analytic procedures used to discover new variants in existing genes and novel genes contributing to LDL-C levels. An important aspect of the result is that no one mechanism can be assumed to be the underlying cause for the association with a complex trait, yet even accumulation of rare variants can have major effects on LDL-C level.
27. Noto D, Cefalù AB, Valenti V, et al. Prevalence of ANGPTL3 and APOB gene mutations in subjects with combined hypolipidemia. Arterioscler Thromb Vasc Biol 2012; 32:805–809.
28. Li B1, Leal SM. Methods for detecting associations with rare variants
for common diseases: application to analysis of sequence data. Am J Hum Genet 2008; 83:311–321.
29. Wu MC, Lee S, Cai T, et al. Rare-variant association testing for sequencing data with the sequence kernel association test. Am J Hum Genet 2011; 89:82–93.
30▪▪. Crosby J, Peloso GM, Auer PL, et al. TG and HDL Working Group of the Exome Sequencing
Project. Loss-of-function mutations in APOC3, triglycerides, and coronary disease. N Engl J Med 2014; 371:22–31.
An example of discovery using exome sequencing that a burden of rare variants in APOC3 is associated with triglyceride levels and that those with the variants have lower triglycerides and lower levels of APOC3. This result was translated to a larger analysis of over 110 000 participants in order to have sufficient numbers of carriers (498) to show that variants had a reduced risk of coronary heart disease.
31. Jorgensen AB, Frikke-Schmidt R, Nordestgaard BG, Tybjaerg-Hansen A. Loss-of-function mutations in APOC3 and risk of ischemic vascular disease. N Engl J Med 2014; 371:32–41.
32▪. Talmud PJ, Shah S, Whittall R, et al. Use of low-density lipoprotein cholesterol gene score to distinguish patients with polygenic and monogenic familial hypercholesterolaemia: a case-control study. Lancet 2013; 381:1293–1301.
A nice application of the gene risk score to examine causality in a meta-analytic framework.
33▪▪. Peloso GM, Auer PL, Bis JC, et al. Association of low-frequency and rare coding-sequence variants with blood lipids and coronary heart disease in 56 000 whites and blacks. Am J Hum Genet 2014; 94:223–232.
A first major use of the ExomeChip to investigate variation in lipids and coronary heart disease risk in 56 538 participants from the CHARGE consortium. Four low-frequency variants in ANGPTL8, PAFAH1B2, COL18A1 and PCSK7 had large effects on HDL-C and triglycerides, but none were associated with the risk of coronary heart disease.
34. Holmen OL, Zhang H, Fan Y, et al. Systematic evaluation of coding variation identifies a candidate causal variant in TM6SF2 influencing total cholesterol and myocardial infarction risk. Nat Genet 2014; 46:345–351.
35. Willer CJ, Sanna S, Jackson AU, et al. Newly identified loci that influence lipid concentrations and risk for coronary artery disease. Nat Genet 2008; 40:161–169.
36. Kozlitina J, Smagris E, Stender S, et al. Exome-wide association study identifies a TM6SF2 variant that confers susceptibility to noalcoholic fatty liver disease. Nat Genet 2014; 46:352–356.
37▪. Tachmazidou I, Dedoussis G, Southam L, et al. A rare functional cardioprotective APOC3 variant has risen in frequency in distinct population isolates. Nat Commun 2013; 4:2872.
A useful manuscript describing approaches and use of isolated populations in discovery of novel genes and variants contributing to complex human phenotypes.
38. Pollin TI, Damcott CM, Shen H, et al. A null mutation in human APOC3 confers a favorable plasma lipid profile and apparent cardioprotection. Science 2008; 322:1702–1705.