Interleukin-6 (IL-6) and receptor (IL6-R) gene haplotypes associate with amniotic fluid protein concentrations in preterm birth.

Spontaneous preterm birth (PTB-gestational age <37 weeks) occurs in approximately 450 000 births annually in the United States and is one of the leading causes of neonatal morbidity and mortality. Risk of PTB is affected by complex gene-environment interactions that are not well understood. We examined the PTB candidate gene, Interleukin 6 (IL-6) and its receptor (IL6-R) in both Caucasian (145 PTB and 194 term maternal; 140 PTB and 179 term fetal) and African-American (76 PTB and 191 term maternal; 66 PTB and 183 term fetal) DNA. Eight single nucleotide polymorphisms (SNPs) in IL-6 and 22 SNPs in IL6R were examined for association with IL-6 amniotic fluid (AF) concentrations, as concentration of IL-6 is a hypothesized risk factor. In addition, IL-6 and IL6-R SNPs were analyzed for associations with PTB. Haplotype associations were tested by sliding windows. No strong single marker effects were observed in Caucasians; however, in African-American maternal IL-6R marker rs4553185 associated with PTB (allele P = 4.49 x 10(-3) and genotype P = 0.01). The strongest haplotype associations were observed in IL-6R with IL-6 cytokine concentration as outcome: Caucasian fetal (rs4601580-rs4845618) P = 1.6 x 10(-3) and African-American maternal (rs4601580-rs4845618-rs6687726-rs7549338) P = 2.30 x 10(-3). Significant results converged on three regions in the two genes: in IL-6 markers rs1800797, rs1800796 and rs1800795; in IL-6R markers rs4075015, rs4601580, rs4645618, rs6687726 and rs7549338 and markers rs4845623, rs4537545 and rs4845625. In conclusion, our results suggest that IL-6 AF concentration, in situations of PTB, result from variation in IL-6 and more importantly IL-6R.

Exploring the performance of Multifactor Dimensionality Reduction in large scale SNP studies and in the presence of genetic heterogeneity among epistatic disease models.

BACKGROUND/AIMS: In genetic studies of complex disease a consideration for the investigator is detection of joint effects. The Multifactor Dimensionality Reduction (MDR) algorithm searches for these effects with an exhaustive approach. Previously unknown aspects of MDR performance were the power to detect interactive effects given large numbers of non-model loci or varying degrees of heterogeneity among multiple epistatic disease models. METHODS: To address the performance with many non-model loci, datasets of 500 cases and 500 controls with 100 to 10,000 SNPs were simulated for two-locus models, and one hundred 500-case/500-control datasets with 100 and 500 SNPs were simulated for three-locus models. Multiple levels of locus heterogeneity were simulated in several sample sizes. RESULTS: These results show MDR is robust to locus heterogeneity when the definition of power is not as conservative as in previous simulation studies where all model loci were required to be found by the method. The results also indicate that MDR performance is related more strongly to broad-sense heritability than sample size and is not greatly affected by non-model loci. CONCLUSIONS: A study in which a population with high heritability estimates is sampled predisposes the MDR study to success more than a larger ascertainment in a population with smaller estimates.

CLCNKB-T481S and essential hypertension in a Ghanaian population.

OBJECTIVE: Prior to the discovery of chloride channel Kb with a variant threonine change to serine at position 481 (CLCNKB-T481S) there were no variants or clinical disorders associated with gain-of-function defects in thick ascending limb of the kidney channels or transporters. CLCNKB-T481S is a novel gain-of-function variant that has been associated with essential hypertension. This finding has not been replicated until our current study. In this study, we re-examined CLCNKB-T481S using a large homogenous population from Ghana, and coupled genetic analyses with the functional characterization of this polymorphism using a mammalian expression system. METHODS: We genotyped CLCNKB-T481S in four ethnically defined control populations and a homogenous cohort of normotensive and hypertensive Ghanaians. Functional analysis was performed by whole-cell patch-clamp recording of tsA201 cells (a cell line derived from the human renal cell line, HEK-293) transiently transfected with ClC-Kb and barttin. RESULTS: CLCNKB-T481S was found more commonly in the African and Caucasian-Americans when compared with the Asian and Hispanic American populations, having minor allele frequencies of 0.20, 0.15, 0.06 and 0.01 respectively. Additionally, CLCNKB-T481S was significantly associated with hypertension in Ghanaian males. In stratified logistic regression analysis with Ghanaian males, we observed a significant odds ratio of 3.29 (1.17-9.20 95% confidence interval, P=0.024) in the recessive model (TT versus AT&TT). Unlike previous results obtained in Xenopus oocytes, coexpression of CLCNKB-T481S with the obligatory accessory subunit barttin in tsA201 cells did not generate larger currents than coexpression of the wild-type allele. CONCLUSIONS: We conclude that CLCNKB-T481S is associated with essential hypertension in males within the Ghanaian population; however, further studies are needed to understand its sex and ethnic segregation as well as to identify cellular factors that account for the divergent functional expression of ClC-Kb-T481S plus barttin in Xenopus oocytes and mammalian cells.

Racial disparity in pathophysiologic pathways of preterm birth based on genetic variants.

OBJECTIVE: To study pathophysiologic pathways in spontaneous preterm birth and possibly the racial disparity associating with maternal and fetal genetic variations, using bioinformatics tools. METHODS: A large scale candidate gene association study was performed on 1442 SNPs in 130 genes in a case (preterm birth < 36 weeks) control study (term birth > 37 weeks). Both maternal and fetal DNA from Caucasians (172 cases and 198 controls) and 279 African-Americans (82 cases and 197 controls) were used. A single locus association (genotypic) analysis followed by hierarchical clustering was performed, where clustering was based on p values for significant associations within each race. Using Ingenuity Pathway Analysis (IPA) software, known pathophysiologic pathways in both races were determined. RESULTS: From all SNPs entered into the analysis, the IPA mapped genes to specific disease functions. Gene variants in Caucasians were implicated in disease functions shared with other known disorders; specifically, dermatopathy, inflammation, and hematological disorders. This may reflect abnormal cervical ripening and decidual hemorrhage. In African-Americans inflammatory pathways were the most prevalent. In Caucasians, maternal gene variants showed the most prominent role in disease functions, whereas in African Americans it was fetal variants. The IPA software was used to generate molecular interaction maps that differed between races and also between maternal and fetal genetic variants. CONCLUSION: Differences at the genetic level revealed distinct disease functions and operational pathways in African Americans and Caucasians in spontaneous preterm birth. Differences in maternal and fetal contributions in pregnancy outcome are also different between African Americans and Caucasians. These results present a set of explicit testable hypotheses regarding genetic associations with preterm birth in African Americans and Caucasians.

Spontaneous preterm birth in African Americans is associated with infection and inflammatory response gene variants.

OBJECTIVE: The objective of the study was to study the genetic risk factors of spontaneous preterm birth (PTB) in African Americans. STUDY DESIGN: Case-control analyses were performed using maternal and fetal deoxyribonucleic acid from 279 African American birth events (82 PTB and 197 term) and 1432 single-nucleotide polymorphisms from 130 candidate genes. Single-locus association and haplotype analyses were performed. RESULTS: The most significant associations were in the maternal interleukin (IL)-15 (rs10833, allele P = 2.91 x 10(-4), genotype P = 2.00 x 10(-3)) gene and the fetal IL-2 receptor B (IL-2RB) (rs84460, allele P = 1.37 x 10(-4), genotype P = 6.29 x 10(-4)) gene. The best models for these markers were additive (rs10833, odds ratio [OR], 0.30; 95% confidence interval [CI], 0.14-0.62; P = 1.0 x 10(-3); rs84460, OR, 2.32; 95% CI, 1.47-3.67; P < 1.0 x 10(-3)). The largest number of significant associations was found in genes related to infection and inflammation. There were overall a larger number of significant associations in infants than in mothers. CONCLUSION: These results support a strong role for genes involved in infection and inflammation in the pathogenesis of PTB, particularly IL-12 and IL-12RB, and indicate that in African Americans there may be complementarity of maternal and fetal genetic risks for PTB.

Haplotype diversity in four genes (CLCNKA, CLCNKB, BSND, NEDD4L) involved in renal salt reabsorption.

OBJECTIVE: Differences exist among various populations with regards to hypertension prevalence, severity, progression and response to therapy. Such differences may be due to genetic or environmental factors. We characterized the genetic variation and haplotype diversity of four hypertension candidate genes (CLCNKA, CLCNKB, BSND, NEDD4L) in four different ethnic groups (Caucasian Americans, African-Americans, Han Chinese, and Mexican-Americans). METHODS: We genotyped 42 single nucleotide polymorphisms across the four genes in equal numbers of each ethnically defined population, then tested for linkage disequilibrium, computed allelic and haplotype frequencies, and compared data across the different ethnic groups. RESULTS: We identified significant genotype and allele frequency differences among ethnic groups. The strongest differences were observed between African-American and Mexican-Americans and between Caucasian and Mexican-Americans. In addition, haplotype blocks were defined for BSND, CLCNKA_B and NEDD4L in the four populations examined. Completely mismatched ('yin yang') haplotypes were also observed. We found that the number of inferred halpotypes varied gene to gene and in some instances between the populations for a given gene indicating substantial haplotype diversity. The haplotype diversity among the various ethnic populations observed in our study was greater than that reported in Perlegen database. CONCLUSIONS: Haplotype diversity in hypertension candidate genes has important implications for designing and evaluating candidate gene or genome-wide blood pressure association studies that consider these genes.

Genetic regulation of amniotic fluid TNF-alpha and soluble TNF receptor concentrations affected by race and preterm birth.

Racial disparity in spontaneous preterm birth (PTB) between African Americans and Caucasians in the US is unexplained, but is probably related to differences in amniotic fluid (AF) inflammatory cytokine profiles. Therefore, this study analyzed the association of 34 single nucleotide polymorphisms (SNPs) in TNF-alpha and its receptor genes (TNFR1 and TNFR2) with AF TNF-alpha and soluble TNF receptor (R1 and R2) concentrations in PTB. Samples consisted of African American and Caucasian cases (PTB), and controls (term birth) for which both cytokine, and maternal and fetal genotype data were available. Analyses were performed with genotype, case, and maker-status interaction in the model for log transformed cytokine concentrations. In Caucasians, two interactions between genotype and pregnancy outcome associated with cytokine concentrations, whereas 14 gene variants in African Americans showed interactions with pregnancy outcome, and 13 showed association with genetic markers. In conclusion, cytokine concentrations in African American preterm births can be partially explained by interactions between pregnancy outcome, SNPs and infection. This does not appear to be the case in Caucasians. These findings may be important in understanding disparity in rates of PTB between the two populations.

Patterns of cytokine profiles differ with pregnancy outcome and ethnicity.

BACKGROUND: Preterm birth (PTB) is hypothesized to be an inflammatory response disease. However, no single factor alone is likely to explain PTB risk. It is more probable that coordinated networks of cytokines affect risk. METHODS: Therefore, we examined the relationships between amniotic fluid (AF) cytokines/chemokines and related biomarkers in PTB and normal term deliveries in African Americans and Caucasians. Data were obtained from African American (41 preterm labor and 91 term labor) and Caucasian (105 preterm labor and 100 term labor) pregnant mothers. Pro-inflammatory cytokines and related molecules interleukin (IL)-1, IL-6, IL-8, and tumor necrosis factor- (TNF)-alpha, TNF soluble receptors (sTNFR1 and sTNFR2), and anti-inflammatory cytokine IL-10 that were all previously associated with PTB were studied. Correlations between biomarkers were calculated; differences of correlation coefficients between AF from African American and Caucasian samples in preterm labor and term labor were measured. RESULTS: Multiple differences were observed between African American and Caucasian preterm and term birth groups. In term birth the strongest differences were between pro- and anti-inflammatory correlations, whereas in PTB differences were equally distributed between pro-inflammatory/anti-inflammatory and pro-inflammatory/pro-inflammatory correlations. Three correlation patterns differed significantly between AF from PTB African Americans with and without microbial invasion of the intra-amniotic cavity (MIAC); no differences were observed in Caucasians with MIAC. CONCLUSION: Correlation analyses of cytokine measurements suggest coordinated interplay during pregnancy; significant differences exist between African Americans and Caucasians. Such analyses can serve as a means of understanding risk factors in these populations.

Mitochondrial DNA variant A4917G, smoking and spontaneous preterm birth.

Spontaneous preterm birth (PTB; <37 weeks gestation) is a major risk factor for neonatal morbidity and mortality. The exact cause of PTB is unknown but oxidative stress may play an important role. Genetic studies have recently begun to elucidate the role of genetic variation in PTB but these studies have overlooked the mitochondrial genome/gene(s) as a plausible PTB candidate. In the present study, we sought to document association between nonsynonymous mitochondrial DNA (mtDNA) variants A4917G, G10398A and T4216C and PTB. We performed a case (PTB; <36 weeks gestation)-control (normal term) analysis of these mtDNA markers and examined their potential interaction with smoking in PTB. A sample of 422 pregnant Caucasian women (220 preterm and 202 terms) was examined for association. Haplogroup T marker A4917G was identified as a possible candidate for association with PTB after adjusting for smoking (OR=1.99 [95% CI 0.93-4.24]) as was T4216C (OR=1.63 [95% CI 0.93-2.83]). No significant multi-locus interactions or interactions with other environmental variables were observed. Our data, although preliminary, support the hypothesis that mitochondrial genome polymorphisms may play a significant role in PTB through an interaction with smoking.

Racial disparity in maternal-fetal genetic epistasis in spontaneous preterm birth.

OBJECTIVE: To understand the differences in genetic interactions among tumor necrosis factor-alpha, interleukin-6 and their receptor gene variants between black and white patients in spontaneous preterm birth. STUDY DESIGN: Maternal and fetal DNA (n = 1195) were collected from cases (preterm birth < 36 weeks' gestation; n = 448), controls (> 37 weeks' gestation; n = 747), and genotyped for single nucleotide polymorphisms in tumor necrosis factor-alpha, tumor necrosis factor receptor 1, and tumor necrosis factor receptor 2, interleukin-6, and interleukin-6 receptor loci. Multifactor dimensionality reduction analysis was used to test all single and multilocus combinations for the ability to predict pregnancy outcome. RESULTS: In white patients, multilocus interactions in maternal DNA between single nucleotide polymorphisms at -7227 (interleukin-6), 22,215 (interleuki-6 receptor) and -3448 (tumor necrosis factor-alpha) was predictive of approximately 59.1% (P < .02; odds ratio, 2.3 [95% confidence interval = 1.6-3.4]) of pregnancy outcome. In white fetal DNA and black maternal DNA, no significant interactive models were observed. In black patients, the best epistatic model was in fetal DNA between single nucleotide polymorphisms at 17,691 (tumor necrosis factor-receptor 1) and at -3448 (tumor necrosis factor-alpha) and was predictive of pregnancy outcome 68.3% of the time (P < .01; odds ratio, 5.0 [95% confidence interval = 2.6-9.6]). CONCLUSION: Analyses of multilocus interactions found/associated different models in black and white patients in both maternal and fetal DNA with preterm birth as outcome. Significant maternal-fetal interactions were not detected in either race.

Functional BSND variants in essential hypertension.

BACKGROUND: Defects in the handling of renal salt reabsorption may contribute to interindividual differences in blood-pressure regulation and susceptibility to hypertension. Sodium chloride reabsorption in the thick ascending limb (TAL) is dependent in part on the chloride channel, ClC-Kb (encoded by CLCNKB), and its accessory subunit, barttin (encoded by BSND). METHODS: We investigated genetic variations in BSND in a screening population, and genotyped a homogenous cohort of normotensive and hypertensive Ghanaian subjects, in addition to four ethnically defined control populations. Functional consequences of the identified BSND variants were examined using a heterologous expression system. RESULTS: Three novel, nonsynonymous coding-sequence single-nucleotide polymorphisms were identified (V43I, E255Q, and G284D) in the screening population. BSND-V43I was identified in African American, Asian, and Hispanic subjects, with minor allele frequencies of 0.14, 0.18, and 0.01, respectively, but it was absent in the Caucasian population. BSND-E225Q and BSND-G284D were rare variants. Two of these variants (V43I and G284D) exhibited partial loss-of-function phenotypes when heterologously expressed with ClC-Kb chloride channels in cultured cells. In logistic regression analyses, we observed no association between hypertension and BSND-I43 in our study population. However, we did observe significant deviation from Hardy-Weinberg equilibrium in the normotensive population. CONCLUSIONS: We conclude that BSND-V43I, a common variant conferring partial loss of function, exhibits significant deviation from Hardy-Weinberg equilibrium in the Ghanaian normotensive control population. However, it does not independently confer protection against hypertension.

Angiotensinogen promoter sequence variants in essential hypertension.

BACKGROUND: Essential hypertension is a complex multifactorial disease caused by ill-defined genetic factors. The angiotensinogen (AGT) gene has been implicated as a risk factor in essential hypertension. METHODS: To assess the role of AGT in hypertension, we evaluated two polymorphisms (A-6G and C-20A) in the 5' region of the gene that have been shown to have a role in transcriptional regulation. A total of 463 subjects were studied: 243 African Americans (26 male and 34 female normotensives, 66 male and 117 female hypertensives) and 220 whites (35 male and 60 female normotensives, 55 male and 70 female hypertensives). African American and white subjects were examined individually, as significant differences in allele and genotype frequencies were observed between these two cohorts. RESULTS: White female hypertensives and normotensives differed significantly in genotype frequency at C-20A (P = .02). No other single site comparisons were significantly different between hypertensives and normotensives in either the white or African American samples. Haplotype frequencies in white males also differed significantly between phenotypic classes (P = .05). To evaluate the data further, we assessed all polymorphic sites simultaneously by the examination of multisite interaction and determined the single best genetic model for each population. A model that included both sites and gender correctly predicted hypertension status in the white population 59.1% of the time (P = .039). The model generated for the African American population was not significant. CONCLUSIONS: Our results suggest that a complex set of genetic factors interact with gender to predispose whites to hypertension.

Multilocus interactions at maternal tumor necrosis factor-alpha, tumor necrosis factor receptors, interleukin-6 and interleukin-6 receptor genes predict spontaneous preterm labor in European-American women.

OBJECTIVE: We hypothesize that genetic variations (single nucleotide polymorphisms-SNPs) in the tumor necrosis factor-alpha (TNF-alpha), TNF receptors (TNFRI and TNFRII), interleukin-6 (IL-6) and IL-6 receptor (IL-6R) genes predict high-risk status for spontaneous preterm birth (sPTB) in European-American women. In this study we examine the allelic and genotypic variations and the gene-gene interactions in the TNF-alpha, TNFRs, IL-6, and IL-6R genes in maternal DNA samples by using a case-control model. STUDY DESIGN: Maternal DNA from cases of sPTB after preterm labor (n = 101) and controls (normal term labor and delivery) (n = 321) were genotyped for SNPs in the TNF-alpha (6), TNFRI (6), TNFRII (7), IL-6 (5), and IL-6R (3) loci. SNPs were tested for both allele and genotype differences (cases vs controls) with the use of standard genetic epidemiologic methods. Multilocus interaction was assessed with multifactor dimensionality reduction analysis (MDR) to test all single and multilocus combinations for the ability to predict sPTB. RESULTS: Few significant allelic and genotypic associations were detected between cases and controls in maternal DNA. Single locus analysis documented independent association of SNPs at -7294 (allele and genotype) of TNFRI and 24660 (genotype) TNFRII loci with sPTB. MDR revealed a significant 3 locus model that includes SNPs -3448 of TNF-alpha, -7227 of IL-6, and 33314 of IL-6R. This interactive model allowed the successful prediction of pre- to low-risk genotypes is 3.50 (95% CI 2.52-4.87, P < .001). CONCLUSION: This is the first report to document a multilocus interaction in sPTB that predicts 65.2% of the cases in a European-American sample. Although putatively significant associations with sPTB were seen at a few single locus sites in TNFRI and TNFRII, they were not as predictive as the 3-locus model produced by MDR, suggesting the use of multilocus analyses in gene association studies of complex disease such as sPTB.

Geographic ancestry and markers of preterm birth.

Several biomarkers associated with spontaneous preterm birth (PTB) have been discovered over the last decade. Many of these markers, such as cytokines, are associated with infection and inflammation. As such, these biomarkers represent biologically plausible candidates for assessing those at risk of PTB. However, in the early association studies of biomarker-pregnancy outcome, the geographic ancestry of subjects was not considered. Based on more recent data, it is becoming increasingly evident that these biomarkers, and a universal approach that uses a single biomarker, fail to provide adequate assessment of risk in all subjects. Rather, recent data support the conclusion that some markers associate in subjects of African descent and another nonoverlapping set associates in subjects of European descent. These data indicate that diagnostic or predictive tests will have to use different biomarkers for different sets of subjects. If this is true, it poses severe restrictions on how to predict outcome or perform tests of association, and may make it impossible to determine risk or provide proper intervention. An alternative is presented that, although not yet proven, may make it possible to use a common set of biomarkers and their relationships to assess risk.

Maternal and fetal genetic associations of PTGER3 and PON1 with preterm birth.

OBJECTIVE: The purpose of this study was to identify associations between maternal and fetal genetic variants in candidate genes and spontaneous preterm birth (PTB) in a Norwegian population and to determine the effect size of those associations that corroborate a previous study of PTB. METHODS: DNA from 434 mother-baby dyads (214 cases and 220 controls) collected from the Norwegian Mother and Child Cohort (MoBa) was examined for association between 1,430 single nucleotide polymorphisms in 143 genes and PTB. These results were compared to a previous study on European Americans (EA) from Centennial Women's Hospital in Nashville, TN, USA. Odds ratios for SNPs that corroborated the Cenntennial study were determined on the combined MoBa and Centennial studies. RESULTS: In maternal samples the strongest results that corroborated the Centennial study were in the prostaglandin E receptor 3 gene (PTGER3; rs977214) (combined genotype p = 3x10(-4)). The best model for rs977214 was the AG/GG genotypes relative to the AA genotype and resulted in an OR of 0.55 (95% CI = 0.37-0.82, p = 0.003), indicating a protective effect. In fetal samples the most significant association in the combined data was rs854552 in the paraoxonase 1 gene (PON1) (combined allele p = 8x10(-4)). The best model was the TT genotype relative to the CC/CT genotypes, and resulted in an OR of 1.32 (95% CI = 1.13-1.53, p = 4x10(-4)). CONCLUSIONS: These studies identify single locus associations with preterm birth for both maternal and fetal genotypes in two populations of European ancestry.

Preterm birth in Caucasians is associated with coagulation and inflammation pathway gene variants.

Spontaneous preterm birth (<37 weeks gestation-PTB) occurs in approximately 12% of pregnancies in the United States, and is the largest contributor to neonatal morbidity and mortality. PTB is a complex disease, potentially induced by several etiologic factors from multiple pathophysiologic pathways. To dissect the genetic risk factors of PTB a large-scale high-throughput candidate gene association study was performed examining 1536 SNP in 130 candidate genes from hypothesized PTB pathways. Maternal and fetal DNA from 370 US Caucasian birth-events (172 cases and 198 controls) was examined. Single locus, haplotype, and multi-locus association analyses were performed separately on maternal and fetal data. For maternal data the strongest associations were found in genes in the complement-coagulation pathway related to decidual hemorrhage in PTB. In this pathway 3 of 6 genes examined had SNPs significantly associated with PTB. These include factor V (FV) that was previously associated with PTB, factor VII (FVII), and tissue plasminogen activator (tPA). The single strongest effect was observed in tPA marker rs879293 with a significant allelic (p = 2.30x10(-3)) and genotypic association (p = 2.0x10(-6)) with PTB. The odds ratio (OR) for this SNP was 2.80 [CI 1.77-4.44] for a recessive model. Given that 6 of 8 markers in tPA were statistically significant, sliding window haplotype analyses were performed and revealed an associating 4 marker haplotype in tPA (p = 6.00x10(-3)). The single strongest effect in fetal DNA was observed in the inflammatory pathway at rs17121510 in the interleukin-10 receptor antagonist (IL-10RA) gene for allele (p = 0.01) and genotype (p = 3.34x10(-4)). The OR for the IL-10RA genotypic additive model was 1.92 [CI 1.15-3.19] (p = 2.00x10(-3)). Finally, exploratory multi-locus analyses in the complement and coagulation pathway were performed and revealed a potentially significant interaction between a marker in FV (rs2187952) and FVII (rs3211719) (p<0.001). These results support a role for genes in both the coagulation and inflammation pathways, and potentially different maternal and fetal genetic risks for PTB.

A balanced accuracy function for epistasis modeling in imbalanced datasets using multifactor dimensionality reduction.

Multifactor dimensionality reduction (MDR) was developed as a method for detecting statistical patterns of epistasis. The overall goal of MDR is to change the representation space of the data to make interactions easier to detect. It is well known that machine learning methods may not provide robust models when the class variable (e.g. case-control status) is imbalanced and accuracy is used as the fitness measure. This is because most methods learn patterns that are relevant for the larger of the two classes. The goal of this study was to evaluate three different strategies for improving the power of MDR to detect epistasis in imbalanced datasets. The methods evaluated were: (1) over-sampling that resamples with replacement the smaller class until the data are balanced, (2) under-sampling that randomly removes subjects from the larger class until the data are balanced, and (3) balanced accuracy [(sensitivity+specificity)/2] as the fitness function with and without an adjusted threshold. These three methods were compared using simulated data with two-locus epistatic interactions of varying heritability (0.01, 0.025, 0.05, 0.1, 0.2, 0.3, 0.4) and minor allele frequency (0.2, 0.4) that were embedded in 100 replicate datasets of varying sample sizes (400, 800, 1600). Each dataset was generated with different ratios of cases to controls (1 : 1, 1 : 2, 1 : 4). We found that the balanced accuracy function with an adjusted threshold significantly outperformed both over-sampling and under-sampling and fully recovered the power. These results suggest that balanced accuracy should be used instead of accuracy for the MDR analysis of epistasis in imbalanced datasets.

Data simulation software for whole-genome association and other studies in human genetics.

Genome-wide association studies have become a reality in the study of the genetics of complex disease. This technology provides a wealth of genomic information on patient samples, from which we hope to learn novel biology and detect important genetic and environmental factors for disease processes. Because strategies for analyzing these data have not kept pace with the laboratory methods that generate the data it is unlikely that these advances will immediately lead to an improved understanding of the genetic contribution to common human disease and drug response. Currently, no single analytical method will allow us to extract all information from a whole-genome association study. Thus, many novel methods are being proposed and developed. It will be vital for the success of these new methods, to have the ability to simulate datasets consisting of polymorphisms throughout the genome with realistic linkage disequilibrium patterns. Within these datasets, we can embed genetic models of disease whereby we can evaluate the ability of novel methods to detect these simulated effects. This paper describes a new software package, genomeSIM, for the simulation of large-scale genomic data in population based case-control samples. It allows for single SNP, as well as gene-gene interaction models to be associated with disease risk. We describe the algorithm and demonstrate its utility for future genetic studies of whole-genome association.

Ethnic differences in key candidate genes for spontaneous preterm birth: TNF-alpha and its receptors.

OBJECTIVES: Spontaneous preterm birth (PTB) has a significant ethnic disparity with people of African descent having an almost 2-fold higher incidence than those of European descent in the United States. This disparity may be caused by differences in the distribution of genetic risk factors. The objective of this study is to examine genetic differences between African-Americans and European Americans for single nucleotide polymorphisms (SNPs) in candidate genes for PTB. METHODS: We examined patterns of variation in 19 SNPs in 3 candidate genes for preterm birth: TNF-alpha, TNF-receptor 1 and TNF-receptor 2. Allele, genotype and haplotype frequencies were compared between African-Americans (AA) and European-Americans (EA) in cases and controls separately. Both maternal and fetal genotypes were studied, as it is unclear whether one or both of these are important in the etiology of PTB. RESULTS: The vast majority of the SNPs differed significantly between ethnic groups, although there are only a few suggestive results comparing cases and controls within an ethnic group. For TNF-alpha, four of six SNPs; for TNF-R1, 5/6; and for TNF-R2, 6/7 showed significant differences between ethnic groups in either allele and/or genotype frequency. CONCLUSIONS: Our data demonstrate highly significant genetic differences between ethnic groups in genes that may play a role in the risk of PTB.