SEARCH

SEARCH BY CITATION

Keywords:

  • metabolic syndrome;
  • quantitative trait loci;
  • coronary vascular disease;
  • type 2 diabetes;
  • hypertension

Abstract

  1. Top of page
  2. Abstract
  3. Research Methods and Procedures
  4. Acknowledgement
  5. References

As part of the Hypertension Genetic Epidemiology Network study, genome scans were performed in two ethnicities on the categorical metabolic syndrome (MetS). Genome scans were performed also on the factor scores produced by factor analysis (quantitative MetS). Heritabilities were highest for the obesity-insulin (INS) factor and lowest for blood pressure (BP) and central obesity. Seventeen unique putative quantitative trait loci (QTLs) yielded logarithm of the odds ratio (LOD) scores in excess of 1.7, 8 for blacks and 9 for whites. Important QTL findings in whites included an LOD score of 3.19 on chromosome 15q15 for the BP factor, 3.08 on chromosome 8p23 for the lipids-INS factor, and 3.07 on chromosome 3p26 for the obesity-INS factor. In blacks, after excluding type 2 diabetics, important QTLs were identified, including an LOD score of 2.77 on 13p12 for the obesity-INS factor and 2.63 on chromosome 11q24 for the lipids-INS factor. Categorical MetS had lower results than quantitative MetS. Notably, several loci identified overlap with those identified in other studies for a single or group of traits. The most promising candidate loci on 11q24 for lipids-INS and 13p12 for obesity-INS in blacks, 8p23 for lipids-INS, 14q24 for obesity-INS, and 15q15 for BP in whites warrant further investigation.

The metabolic syndrome (MetS)1 is expressed in humans as a cluster of metabolic abnormalities that include obesity, insulin (INS) resistance, dyslipidemia, and hypertension. A higher prevalence of MetS in a population is often associated with higher risk for coronary vascular disease and type 2 diabetes (T2D). Employing the National Cholesterol Education Program (NCEP) definition, we recently found a higher prevalence of categorical MetS (c-MetS) in the Hypertension Genetic Epidemiology Network (HyperGEN) blacks and whites than in the general U.S. population (1). Details of these analyses are reported elsewhere (2). We use the term quantitative MetS (q-MetS) for the latent factors created from the multivariate factor analyses (FAs) of 11 MetS risk variables. Details of our analyses on q-MetS are reported elsewhere (2).

Several epidemiological and experimental studies have examined c-MetS and q-MetS (2, 3, 4, 5, 6, 7, 8). Only a few have investigated the genetic nature of MetS (9, 10, 11, 12, 13, 14). Identification of genomic regions harboring genes predisposing to the MetS is the main purpose of this article. This report is based on data from the HyperGEN study described in detail by Williams et al. (15). We present a non-parametric genome-wide linkage analysis of c-MetS and variance components linkage analysis for four factor domains inferred for q-MetS with and without Varimax rotation in the HyperGEN study (2).

Heritability estimates suggested excellent prospects for the identification of quantitative trait loci (QTLs). In blacks, the heritabilities were evaluated for the factors obesity-INS, blood pressure (BP), lipids-INS, and central obesity in the respective ranges 50% to 60%, 34% to 38%, 28% to 34%, and 28% to 46%. In whites, the respective heritability ranges were 45% to 65%, 33% to 37%, 40% to 46%, and 28% to 37%.

Table 1 summarizes all logarithm of the odds ratio (LOD) scores above 1.7. The results of q-MetS without rotation mainly repeated the most important findings from the q-MetS with rotation. Consequently, they will only be discussed in conjunction with Table 1 results. Results for c-MetS were less interesting, and we will describe only a few of them.

Table 1. . Linkage analysis LOD score (LOD > 1.7) results (rotation Varimax)
Ethnic groupData typeChromosomeMarker (Marshfield)D numberSex-averaged distance (cM)Factor domainLOD score
  • *

    Cytogenetic location encompasses the marker mentioned on the same row of the table.

BlacksAll2p25GATA116B01D2S295217.88Central obesity1.96
  7q22GATA5D08D7S821109.12Obesity-INS1.73
  13p12GATA23C03D13S7878.87Obesity-INS1.77
  17q25AFM044XG3D17S784116.86Central obesity1.90
 All Excluding T2D3q13.3GATA68F07D3S2460134.64Lipids-INS1.86
  3q13.3ATA34G06D3S4523138.00Lipids-INS1.85
  7q21GATA3F01D7S82098.44Obesity-INS2.11
  11q24AFM157XH6D11S912131.26Lipids-INS2.63
  13p12GATA23C03D13S7878.87Obesity-INS2.77
  13p12GATA23C03D13S7878.87Lipids-INS1.93
  14q13ATA29G03D14S59940.68Obesity-INS1.81
WhitesAll1p36.3AFM280WE5D1S4684.22Obesity-INS2.28
  3p26GATA22G12D3S23875.54Obesity-INS3.07
  4q21GATA10G07D4S324388.35BP2.03
  8p23AFM143XD8D8S2640.73Lipids-INS3.08
  10q23GATA64A09D10S1239125.41Central obesity1.87
  15q15GATA151F03D15S150760.17BP3.17
 All Excluding T2D1p36.3AFM280WE5D1S4684.22Obesity-INS1.73
  1q43-44GATA50F11D1S1609274.53Obesity-INS2.07
  3p26GATA22G12D3S23875.54Obesity-INS2.42
  3p24GATA164B08D3S454526.25Obesity-INS1.84
  4q21GATA10G07D4S324388.35BP1.81
  8p23AFM143XD8D8S2640.73Lipids-INS2.31
  14q24GGAA21G11D14S617105.53Obesity-INS2.44
  15q15GATA151F03D15S150760.17BP3.19

Of particular interest were five genomic regions, each with an LOD score over 3. In whites, the BP factor yielded an LOD score of 3.17 (2.84 without rotation) in all data and 3.19 (2.92 without rotation) when excluding T2D on chromosome 15q15 at 60.17 centimorgans (cM); the Obesity-INS factor yielded an LOD score of 3.07 (2.32 without rotation for central obesity) on chromosome 3p26 at 5.54 cM, which was reduced to 2.42 (2.21 without rotation for central obesity) when excluding T2D; the obesity-INS factor yielded an LOD score of 2.44 on chromosome 14q24 at 105.53 cM (2.44 at 105.53 cM and 3.07 at 113.17 cM without rotation) after excluding type 2 diabetics; and the lipids-INS factor yielded an LOD score of 3.08 (2.49 without rotation) on chromosome 8p23 at 0.73 cM (2.31 when excluding T2D) (Table 1 and Figure 2; figures are available online at the Obesity Research web site, www.obesityresearch.org).

In blacks, the strongest evidence was found for the obesity-INS factor on chromosome 13p12 at 8.87 cM, with an LOD score of 1.77 (1.84 without rotation) in all data, which increased to 2.77 (3.54 without rotation at the same location and 2.69 at 25.8 cM) when excluding T2D participants (Table 1 and Figure 1; figures are available online at the Obesity Research web site, www.obesityresearch.org). Additional promising regions were identified, although with smaller LOD scores (Table 1).

Linkage analyses performed in multiple groups can pose a serious problem related to multiple comparisons. However, the results of all data, removing type 2 diabetics, and with or without Varimax rotation replicated the primary findings. On the other hand, analyzing MetS as a categorical trait reduced the power for gene discovery compared with the results produced by genomic scans on the MetS factors. Non-parametric linkage analysis of the affection status produced only three peaks in whites that passed a threshold of 1.5 LOD score (1.53 on chromosome 3 at 69.29 cM, 1.64 on chromosome 3 at 101.32 cM, and 1.76 on chromosome 17 at 31.33 cM).

Table 2 summarizes results from other studies that overlap with our findings. These findings for phenotypes that contribute to a MetS factor are reproduced here for three reasons: they serve as evidence for replications of the HyperGEN MetS QTLs, they can be seen as pointing to a cluster location for genes for other MetS-related traits, and one or more genes with pleiotropic effects may influence the QTLs reported for different traits and MetS. For example, Naumova et al. and An et al. (see Table 2) reported suggestive linkage for familial combined hyperlipidemia and INS, respectively, close to the HyperGEN lipids-INS QTL location on chromosome 8. Also of interest among our findings is the peak for the BP factor on chromosome 15q15, location 60.17 cM, with an LOD 3.19. At this location, Arya et al. (11) reported a BP LOD of 2.0, and Almasy et al. (see Table 2) reported a high-density lipoprotein (HDL) cholesterol LOD of 3.26 located at 62-cM Marshfield map position. Watanabe at al. (see Table 2) reported an INS LOD of 2.67 close to 14q24 where an obesity-INS QTL for whites is located.

Table 2. . LOD score results as additional support from other studies
ChromosomeLOD/p valueReported marker(s)Distance Marshfield sex average (cM)PhenotypeSample characteristicsEthnicityReferences
  Original location (cM)     
  • *

    NPL method.

  • NA, not available.

  • FCHL, familial combined hyperlipidemia, cholesterol, and TG level ≥ age-sex-specific 95th percentile.

32.100D3S1271 TG/HDLC535Indo-MauritianFrancke et al., Hum Mol Genet. 2001;10(24):2751–65
  130.45117.76 T2D and CHD  
41.570D4S2964 LDLC/HDLC535Indo-MauritianFrancke et al., Hum Mol Genet. 2001;10(24):2751–65
  92.0788.35 T2D and CHD  
42.520*D4S2361 FCHL lipid503WhitesNaumova et al., Arterioscler Thromb Vasc Biol. 2003;23:2070–77
  93.8093.48 High CHR and TG  
73.200D7S479/D7S471 BP261HispanicsArya et al., Diabetes 2002;51:841–47
  130.00109.12–123.01 Non-diabetics  
72.05D7S506/D7S653 TG418HispanicsDuggirala et al., Am J Hum Genet. 2000;66:1237–45
  94.0073.84–84.52 Familial combined hyperlipidemia  
71.940NA INS1175WhitesWatanabe et al., Am J Hum Genet. 2000;67:1186–200
  101.00NA Diabetics  
82.200*D8S264 FCHL lipid503WhitesNaumova et al., Arterioscler Thromb Vasc Biol. 2003;23:2070–77
  0.000.73 High CHR and TG  
82.230D8S277 INS322 sib pairsWhitesAn et al., Metabolism. 2003;52(2):246–53
  4.058.34 Sedentary families  
113.60D11S4464/D11S912 BMI1664 sib pairsPima IndiansHanson et al., Am J Hum Genet. 1998;63(4):1130–38
  105–139123.00–131.26 Selected for T2D  
111.70D11S4464/D11S912 T2D1766 sib pairsPima IndiansHanson et al., Am J Hum Genet. 1998;63(4):1130–38
  105–139123.00–131.26 Selected for T2D and obesity  
110.004D11S934 BP263WhitesSharma et al., Hypertension. 2000;35:1291–96
  126.21126.21 Affected sibs  
112.36157xh6 BMI2226WhitesStrug et al., BMC Genetics. 2003;4(Suppl 1):S14
  131.00NA The Framingham Heart Study  
111.18NA TG622BlacksCoon et al., Arterioscler Thromb Vasc Biol. 2001;21:1969–76
  127.50127.50 Ascertained for hypertension  
133.280D13S175/D13S221 BMI1175WhitesWatanabe et al., Am J Hum Genet. 2000;67:1186–200
  5.006.03–12.91 Diabetics  
133.200D13S257 BMI2450WhitesFeitosa et al., Am J Hum Genet. 2002;70:72–82
  42.00NA CHD, high risk scores for CHD  
135.700D13S257 BMI834WhitesBorecki et al., Abstracts. 2002; IGES-16:269
  42.00NA Subgroup of FHS data  
132.01D13S171/D13S263 HDLC379WhitesElbein and Hasstedt, Diabetes. 2002;51:528–35
  42.0025.08–38.32 At least one T2D in a family  
142.670D14S95 INS1175WhitesWatanabe et al., Am J Hum Genet. 2000;67:1186–200
  102.00NA Diabetics  
152.000D15S549/D15S103 BP261HispanicsArya et al., Diabetes 2002;51:841–47
  61.0043.47–NA Non-diabetics  
153.260D15S643 HDLC477HispanicsAlmasy et al., Am J Hum Genet. 1999;84:1686–93
  62.0052.33 Randomly ascertained  

In blacks, the strongest evidence was found for the obesity-INS factor on chromosome 13p12 at 8.87 cM (Table 1 and Figure 1; figures are available online at the Obesity Research web site, www.obesityresearch.org). Watanabe et al. reported a BMI LOD of 3.28 at a close location. From our data under the same peak, an LOD score of 1.93 was found for lipids-INS. Other pertinent findings that overlapped with our q-MetS QTLs were for INS (Watanabe et al.), triglycerides (TGs) (Duggirala et al.), and BP (11) on chromosome 7; for BP (Sharma et al.), BMI and T2D (Hanson et al.) and BMI (Strug et al.) and TG (Coon et al.) on chromosome 11 (see Table 2).

This article also includes additional findings of interest, although less prominent. An example relates to chromosome 13 in the whites after excluding T2D participants. Central obesity factor yielded an LOD score of 1.64 located at 45.55 cM at marker GATA29A09 (D13S788). Feitosa et al. and Borecki et al. (see Table 2) reported BMI LOD scores of 3.2 and 5.7, respectively, at 42.00 cM. Elbein and Hasstedt (see Table 2) reported an HDL LOD of 2.01 on the same general location. Thus, our study results overlap several other published results and, therefore, create premises for finer future studies in the identified regions.

Although the factor scores domains constitute a statistical construct, their structure is not unique only to the HyperGEN data. They represent a latent relationship among important original risk factors and emerge directly from the data structures. A MetS structure that includes obesity, INS resistance, dyslipidemia, and elevated BP was recognized also by the National Heart, Lung, and Blood Institute (NHLBI)/American Heart Association/American Diabetes Association Conference in 2004. Therefore, our findings are applicable to the specific MetS domains identified. Conversely, a potential limitation of our findings is the study design, based on recruiting sibships with at least two hypertensives, but it is difficult to imagine how this specific design feature could lead to false positives, which would be the major worry for any linkage finding.

In conclusion, the most significant QTLs located on chromosomes 3p26, 13p12, and 14q24 for obesity-INS, 11q24 and 8p23 for lipids-INS, and 15q15 for BP factor are promising findings for follow-up studies.

Research Methods and Procedures

  1. Top of page
  2. Abstract
  3. Research Methods and Procedures
  4. Acknowledgement
  5. References

Participants

A total of 4781 participants, contributing at least one of five categories of the NCEP MetS definition, represented the c-MetS sample. Of this, 771 for blacks and 952 for whites were considered affected (having at least three categories beyond the NCEP thresholds). Subsequent to excluding individuals with missing data for any of the 11 risk factors, complete data for analyzing q-MetS were available on 1422 blacks (1173 after excluding T2D) and 1470 whites (1322 after excluding T2D).

MetS

A categorical variable for MetS was created, where individuals were considered affected only if they passed three or more of the five NCEP thresholds for MetS. The NCEP thresholds were: body waist, >102 cm in men and >88 cm in women; fasting TG ≥ 150 mg/dL; HDL, <40 mg/dL in men and <50 mg/dL in women; systolic BP ≥ 130 mm Hg and/or diastolic BP ≥ 85 mm Hg or use of antihypertensive medications; and fasting glucose (GLUC) ≥ 110 mg/dL or use of hypoglycemic medications. In the c-MetS definition, we considered individuals on antihypertensive medications as having met the BP criterion and individuals on antiglycemic medications as having met the GLUC criterion. In all, 33.52% of the blacks and 38.53% of the whites had c-MetS based on the NCEP definition (1, 2). FA was performed to generate factor scores (described in detail elsewhere in (2)) to quantitatively assess any putative QTLs for MetS. The following variables were included in the FA: BMI, kilograms per meter squared; body waist, centimeters; waist-to-hip ratio; GLUC, milligrams per deciliter; fasting INS, microunits per milliliter; systolic BP, milligrams of mercury; diastolic BP, milligrams of mercury; low-density lipoprotein cholesterol, milligrams per deciliter; HDL, milligrams per deciliter; TG, milligrams per deciliter; and percentage body fat (%BF) derived from bioelectric impedance according to the Lukaski formula (16).

Factors scores, from the FA with and without Varimax rotation, were derived separately for blacks and whites once using all of the data and again by excluding type 2 diabetics. In this study, T2D was defined as either having fasting plasma GLUC levels ≥126 mg/dL or by the reported use of INS or antidiabetic medications or a self-report of diabetes diagnosed by a physician with an age of onset ≥ 40 years (1, 5). HyperGEN did not recruit any type 1 diabetics.

Adjustment of the Variables

TG, INS, and HDL were natural log transformed. GLUC and %BF were transformed to the inverse of the square for GLUC and square for %BF. All of the transformed variables fell into a normal distribution. Each of the 11 variables was then adjusted within race and gender groups for the effects of age, age2, age3, and field center. Center variables were created as categorical variables with values of 0 and 1 (absence/presence of a center). One center variable was created for blacks (two field centers) and three for whites (four field centers). Each adjusted normal variable was standardized to mean of 0 and variance of 1.

Genotyping and Linkage Analysis

The NHLBI Mammalian Genotyping Service, Marshfield, performed genotyping with 391 microsatellite markers. Marker quality control was achieved with different software (17, 18, 19). Genome-wide linkage analyses were performed on a total of 370 autosomal markers. The Marshfield human genetic linkage map was used for identifying marker chromosomal distances. Allele frequencies were calculated based on race-specific random samples recruited in HyperGEN. A non-parametric linkage analysis of the affection status (c-MetS) was performed using GENEHUNTER software (18). For the q-MetS, the larger families were split into nuclear families in advance. Multipoint identity by descent probability coefficients was estimated with MAPMAKER/SIBS software at the marker locations (18). Heritability coefficients for each factor domain were estimated by the maximum likelihood method in SEGPATH leaving out of the model the extra spouse correlation. Linkage analysis was performed on all four factors of FA with and without Varimax rotation, derived before and after excluding the type 2 diabetics, using the variance components method as implemented in the SEGPATH software (20). The underlying model postulated an additive effect of a QTL (g), a residual polygenic component (GR), and a residual non-familial variance component (r).

Acknowledgement

  1. Top of page
  2. Abstract
  3. Research Methods and Procedures
  4. Acknowledgement
  5. References

We are thankful to the study participants and the primary centers and investigators of HyperGEN: University of Utah (Network Coordinating Center, Field Center, and Molecular Genetics Laboratory), Steven C. Hunt, (Network Director and Field Center Principal Investigator), Mark F. Leppert (Molecular Genetics P.I.), Jean-Marc Lalouel, Robert B. Weiss, Paul N. Hopkins, Hilary Coon, Roger R. Williams (late), and Janet Hood; Boston University (Field Center), R. Curtis Ellison (P.I.), Richard H. Myers, Luc Djoussé, Yuqing Zhang, Jemma B. Wilk, and Greta Lee Splansky; University of Alabama (Field Center), Albert Oberman (P.I.), Cora E. Lewis, Michael T. Weaver, Phillip Johnson, Susan Walker, and Christie Oden; University of Minnesota (Field Center and Biochemistry Laboratory), Donna K. Arnett (Field Center P.I.), John H. Eckfeldt (Biochemistry Laboratory P.I.), James S. Pankow, Michael B. Miller, Anthony A. Killeen, Kim Weis, Greg Rynders, Catherine Leiendecker-Foster, Gregory Feitl, Barbara Lux, and Jean Bucksa; University of North Carolina (Field Center), Gerardo Heiss (P.I), Barry I. Freedman, Kari E. North, Kathryn Rose, and Amy Haire; Washington University (Data Coordinating Center), D.C. Rao (P.I.), Michael A. Province, Aldi Kraja, Ingrid B. Borecki, Charles Gu, Treva Rice, Mary Feitosa, Jun Wu, Karen L. Schwander, Derek Morgan, Stephen Mandel; Shiping Wang M.S.; Matthew Brown; and NHLBI, Susan E. Old, Cashell Jaquish, and Dina Paltoo. We thank the anonymous reviewers for the constructive suggestions in improving the article.

Footnotes
  • 1

    Nonstandard abbreviations: MetS, metabolic syndrome; INS, insulin; T2D, type 2 diabetes; NCEP, National Cholesterol Education Program; c-MetS, categorical MetS; HyperGEN, Hypertension Genetic Epidemiology Network; q-MetS, quantitative MetS; FA, factor analysis; QTL, quantitative trait locus; BP, blood pressure; LOD, logarithm of the odds ratio; cM, centimorgan(s); HDL, high-density lipoprotein; TG, triglyceride; NHLBI, National Heart, Lung, and Blood Institute; GLUC, glucose; %BF, percentage body fat.

References

  1. Top of page
  2. Abstract
  3. Research Methods and Procedures
  4. Acknowledgement
  5. References
  • 1
    Third Report of the National Cholesterol Education Program (NCEP) (2002) Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel). Final report. Circulation. 37/3: 31443121.
  • 2
    Kraja, A. T., Hunt, S. C., Pankow, J. S., et al (2005) An evaluation of the metabolic syndrome in the HyperGEN study. Nutr Metab (Lond). 2: 2
  • 3
    Meigs, JB. (2000) Invited commentary: insulin resistance syndrome? Syndrome X? Multiple metabolic syndrome? A syndrome at all? Factor analysis reveals patterns in the fabric of correlated metabolic risk factors. Am J Epidemiol. 152: 908912.
  • 4
    Shen, B., Todaro, J. F., Niaura, R., et al (2003) Are metabolic risk factors one unified syndrome? Modeling the structure of the metabolic syndrome X. Am J Epidemiol. 157: 701711.
  • 5
    Hunt, K. J., Heiss, G., Sholinsky, P. D., Province, M. A., FHS Investigators (2000) Familial history of metabolic disorders and the multiple metabolic syndrome: The NHLBI Family Heart Study. Genet Epi. 19: 395409.
  • 6
    Reaven, G. (2002) Metabolic syndrome: pathophysiology and implications for management of cardiovascular disease. Circulation. 106: 286288.
  • 7
    Reilly, M. P., Rader, DJ. (2003) The metabolic syndrome: more than the sum of its parts? Circulation. 108: 15461551.
  • 8
    Zimmet, P., Thomas, CR. (2003) Genotype, obesity and cardiovascular disease: has technical and social advancement outstripped evolution? JInt Med. 254: 114125.
  • 9
    Kissebah, A. H., Sonnenberg, G. E., Myklebust, J., et al (2000) Quantitative trait loci on chromosomes 3 and 17 influence phenotypes of the metabolic syndrome. Proc Natl Acad Sci USA. 97: 1447814483.
  • 10
    Francke, S., Manraj, M., Lacquemant, C., et al (2001) A genome-wide scan for coronary heart disease suggests in Indo-Mauritanians a susceptibility locus on chromosome 16p13 and replicates linkage with the metabolic syndrome on 3q27. Hum Mol Genet. 10: 27512765.
  • 11
    Arya, R., Blangero, J., Williams, K., et al (2002) Factors of insulin resistance syndrome-related phenotypes are linked to genetic locations on chromosomes 6 and 7 in non-diabetic Mexican-Americans. Diabetes. 51: 841847.
  • 12
    Goldin, L. R., Camp, N. J., Keen, K. J., et al (2003) Analysis of metabolic syndrome phenotypes in Framingham Heart study families from Genetic Analysis Workshop 13. Genet Epi. 25(Suppl 1): S78S89.
  • 13
    Loos, RJF, Katzmarazyk, P. T., Rao, D. C., et al (2003) Genome-wide linkage scan for the metabolic syndrome in the Heritage Family Study. J Clin Endocrinal Metab. 88: 59355943.
  • 14
    Tang, W., Miller, M. B., Rich, S. S., et al (2003) Linkage analysis of a composite factor for the multiple metabolic syndrome. Diabetes. 52: 28402847.
  • 15
    Williams, R. R., Rao, D. C., Ellison, R. C., et al (2000) NHLBI family blood pressure program: methodology and recruitment in the HyperGEN network: hypertension genetic epidemiology network. Ann Epidemiol. 10: 389400.
  • 16
    Lukaski, HC. (1989) Use of bioelectric impedance analysis to assess human body composition: a review. In: Levingston GE, ed. Nutritional Status Assessment of the Individual Food and Nutrition Press Inc.: Trumball, CT.
  • 17
    Abecasis, G. R., Cherny, S. S., Cookson, WOC, Cardon, LR. (2001) GRR: graphical representation of relationship errors. Bioinformatics. 17: 742743.
  • 18
    Kruglyak, L., Daly, M. J., Reeve-Daly, M. P., Lander, ES. (1996) Parametric and non-parametric linkage analysis: a unified multipoint approach. Am J Hum Genet. 58: 13471363.
  • 19
    O'Connell, J., Weeks, DE. (1998) Pedcheck: a program for identification of genotype incompatibilities in linkage analysis. Am J Hum Genet. 63: 259266.
  • 20
    Province, M. A., Rice, T. K., Borecki, I. B., Gu, C., Kraja, A., Rao, DC. (2003) Multivariate and multilocus variance components method, based on structural relationships to asses quantitative trait linkage via SEGPATH. Genet Epi. 24: 128138.