Associations between SNPs in candidate immune-relevant genes and rubella antibody levels: a multigenic assessment
© Pankratz et al; licensee BioMed Central Ltd. 2010
Received: 20 May 2010
Accepted: 5 October 2010
Published: 5 October 2010
The mechanisms of immune response are structured within a highly complex regulatory system. Genetic associations with variation in the immune response to rubella vaccine have typically been assessed one locus at a time. We simultaneously assessed the associations between 726 SNPs tagging 84 candidate immune response genes and rubella-specific antibody levels. Blood samples were obtained from 714 school-aged children who had received two doses of MMR vaccine. Associations between rubella-specific antibody levels and 726 candidate tagSNPs were assessed both one SNP at a time and in a variety of multigenic analyses.
Single-SNP assessments identified 4 SNPs that appeared to be univariately associated with rubella antibody levels: rs2844482 (p = 0.0002) and rs2857708 (p = 0.001) in the 5'UTR of the LTA gene, rs7801617 in the 5'UTR of the IL6 gene (p = 0.0005), and rs4787947 in the 5'UTR of the IL4R gene (p = 0.002). While there was not significant evidence in favor of epistatic genetic associations among the candidate SNPs, multigenic analyses identified 29 SNPs significantly associated with rubella antibody levels when selected as a group (p = 0.017). This collection of SNPs included not only those that were significant univariately, but others that would not have been identified if only considered in isolation from the other SNPs.
For the first time, multigenic assessment of associations between candidate SNPs and rubella antibody levels identified a broad number of genetic associations that would not have been deemed important univariately. It is important to consider approaches like those applied here in order to better understand the full genetic complexity of response to vaccination.
The importance of developing protective humoral immunity following vaccination is widely recognized, as those who fail to respond are at increased risk of contracting the disease if exposed. Rubella is well controlled via vaccination programs in industrialized countries, but epidemics of the disease occasionally occur in developing countries and both rubella virus infection and congenital rubella syndrome remain a major health concern around the world [1, 2]. Understanding how host genetic influences modify response to rubella immunization may shed light into the biology of immunity to rubella infection, as well as into the potential development of even more highly effective vaccines. While the heritability of antibody responses to rubella vaccination has been estimated to be as high as 46% , knowledge of the genetic control of rubella vaccine-induced immunity remains incomplete.
Our group and others have shown that polymorphisms in the human leukocyte antigen (HLA) region, as well as SNPs in cytokine and cytokine receptor genes, are associated with differences in a variety of immune responses to rubella vaccine, but do not explain all of the variance in immune responses seen within the population [4–16]. Studies with other viral vaccines, such as measles and mumps, have demonstrated associations between cytokine and cytokine receptor gene polymorphisms and immune responses [17, 18]. Because of the central role of cytokines as intercellular protein messengers and the role of their receptors in the immune response cascade, cytokine and cytokine receptor gene polymorphisms may significantly influence the outcome of rubella vaccine immune response. For example, polymorphisms in both coding and noncoding regions of these genes can affect multiple aspects of cytokine biology, such as transcriptional activity protein production, receptor binding and functional activity [19, 20]. Thus, a wide variety of genes is likely to be important in regulating immune response to live viral vaccines.
While the role of cytokines in antiviral immune responses has been established, little is known about how other gene families control immune responses to rubella virus. Studies in a variety of other models (viruses, bacteria, microbial antigens) have also recently demonstrated the importance of innate and vitamin receptor genes in regulating immune responses [21–23]. In this regard, innate antiviral factor TRIM, toll-like receptors (TLR) and their associated intracellular signaling molecules activation is critical to stimulating innate and adaptive immunity . Importantly, innate pathways detect infection and serve two purposes: mediate initial anti-viral response and prime more powerful and specific adaptive responses. Finally, vitamins and their receptors are known to have hormone-like attributes and were also found to affect innate and adaptive immunity [11, 25]. To further characterize the impact of immune gene polymorphisms on variability in vaccine-induced humoral immunity, it is crucial to broadly examine variants in key genes important to the immune response to viral vaccines such as rubella. The genetic diversity of innate, adaptive, antiviral effector and other immune response genes has not been comprehensively studied within the context of rubella vaccine-induced humoral immune responses.
The discovery of genetic variations caused by single nucleotide polymorphisms (SNPs) has led to population-based immunogenetic studies intended to elucidate the potential relationship between host genomic variation and immune response . A high level of regulatory complexity is required in the human immune system to insure a high probability of functional redundancy in both cell-mediated and humoral immune responses to vaccination [27, 28]. For instance, one gene may be able to compensate for potential loss of function due to genetic variation in another. This would lead to reduced power to detect real associations because variation in immune response may only be apparent within subjects with genetic variants in both of the genes. Because such associations could be missed with single SNP analyses, when associations between genetic variants and measures of immune response are studied it is important to examine more than just their relationships with individual SNPs. That is, not only should genetic associations be studied for the main effects of single genetic variants on immune response, but multigenic associations, whether epistatic or otherwise, should be examined as well [29, 30]. A growing collection of approaches is becoming available for the study of multigenic associations in association with disease states. These include relatively standard statistical approaches to study interactions, but also newer approaches that have been designed to search for the presence of multigenic effects in high dimensional genetic data consisting of many SNPs [31–38].
In this study, we searched for evidence of multigenic associations among a broad collection of SNPs identified from non-HLA genes that are candidates for being immunologically relevant to the development of a humoral response to rubella vaccination after two doses of measles-mumps-rubella vaccine. Due to the large number of genes in immune response gene families, we selected a group of key genes for each family, which, based on biologic plausibility and the published literature, are highly likely to be involved in regulating rubella immunity. We were particularly interested in assessing the degree to which the inclusion of multilocus associations would identify important genetic variation that contributes to rubella-specific antibody levels. To date, no such data have been published to address this issue.
Characteristics of the study population and their associations with rubella-specific IgG antibody levels (reported in IU/ml).
Number of subjects
Q1 - Q3a
19.2 - 63.7
Age at enrollment (years)
19.1 - 67.1
19.4 - 58.5
19.2 - 70.2
19.2 - 57.9
Age at first rubella vaccination (months)
19.6 - 59.9
19.0 - 62.2
18.9 - 72.6
21.8 - 66.2
Age at second rubella vaccination (years)
16.6 - 55.3
20.5 - 69.6
18.9 - 65.4
22.1 - 69.7
20.0 - 69.7
18.3 - 56.7
17.7 - 62.9
19.2 - 63.7
SNP associations identified through the model selection procedure, sorted from the smallest to the largest multi-SNP p-value.
Median (Q1 - Q3)a
0 Minor Alleles
1 Minor Allele
2 Minor Alleles
As with the assessment of all possible pair-wise interactions, the use of recursive partitioning algorithms did not yield strong evidence in favor of major epistatic interactions. After forming the recursive partitioning tree, an assessment of the degree of pruning required to avoid over-fitting indicated that the best-fitting tree was one with a single node, with the split being on rs7801617. Because of this finding, further analyses seeking additional multi-allelic epistatic effects were not pursued.
Because of the regulatory complexity of immune responses to complex antigens, there is almost certainly a diversity of genes that together influence the immune response to vaccination. While there are many genes that are strong candidates, the bulk of those that have an effect on the levels of rubella-specific antibody raised in response to immunization have yet to be identified. In this study, we examined a broad collection of SNPs that tag 84 genes that have been identified as potentially playing a role in the immune process. We have performed an array of analyses in an attempt to identify potential multigenic associations between these candidate SNPs and the observed levels of rubella-specific antibody levels following MMR-II vaccination.
As we have reported previously, a number of SNPs and HLA alleles have already been identified as being univariately associated with variation in responses to rubella immunization [7–16]. In the analyses presented here, we were able to examine all univariate SNP associations with rubella antibodies simultaneously. This assessment suggested that the top four SNPs were among those that were significantly associated with the humoral immune response. When broader analyses were performed that explicitly searched for multiple SNPs that independently contributed to variation in rubella antibody levels while statistically correcting for multiple testing using a randomization approach, we found that there was a significant association between the number of positively associated alleles carried by an individual across all candidate SNPs and differences in rubella antibody level (p = 0.029). In order to identify the subset of SNPs that captured this signal, we subsequently performed a SNP selection procedure and identified a total of 29 SNPs that simultaneously provided information on the levels of rubella-specific antibodies (p = 0.017). In both of these analyses, we employed resampling procedures to account for the process of model building and selection as the tests for significance were performed .
We performed two separate analyses in our search for epistatic effects among the 726 SNPs. We first assessed the significance of all possible two-way combinations of SNPs (263,175 statistical tests). While there were more SNP pairs with p-values between 0.01 and 0.0001 than would have been expected under the null hypothesis, we found that the smallest p-value was actually larger than what would be expected under the hypothesis that no SNP pairs were significantly associated with rubella antibody levels. Likewise, as we performed a recursive partitioning analysis, we found that the best-fitting classification tree had only one node. This provided further evidence against the presence of major epistatic associations being in effect among the SNPs from the candidate genes under study.
While the evidence for epistasis among the candidate SNPs is not strong, our analysis supports the concept that considering the broader collection of a large number of SNPs does provide additional insight into the genetic control of rubella antibody levels. By assessing all SNPs in a single model we were able to identify a broader collection of candidates than was possible when each SNP was considered individually. While the univariate tests implicated SNPs in the LTA, IL6 and IL4R genes, the analysis of all SNPs identified not only these genes but also a broader collection of genes that appear to contribute to variation in rubella antibody levels. The 29 SNPs that were jointly associated with humoral immunity are shown in Table 2. These SNPs that were jointly associated with differences in rubella antibody levels resided in a total of 23 genes.
While there is evidence in favor of there being a major genetic component to the observed levels of rubella-specific antibodies following vaccination , the current knowledge of the genetic control of immune responses to rubella immunization has been obtained from studies that focused on a small number of genetic loci that were analyzed one at a time [7–9, 11–14]. Given the genetic diversity required by the immune system to mount an adequate response to a wide number of pathogens, it is likely that these one-at-a-time approaches are inadequate for the comprehensive study of the genetic correlates of immune response. Indeed, the efforts reported here implicated a broader number of candidate genes as potentially being involved in modulating rubella-specific antibody levels using multigenic analyses that examined more than one locus at a time.
The genes that have been implicated as being associated with rubella antibody levels were selected for study because of their general relevance to viral immunity. The humoral immune response to rubella vaccine reflected in assayed antibody levels is the cumulative result of the actions and interactions of multiple genes and pathways. Those SNPs that were identified as being important for humoral immunity to rubella are included in Table 2. Among these genes are cytokine (Th1/Th2/inflammatory) and cytokine receptor, lymphotoxin alpha (LTA), leukocyte specific transcript-1 (LST1), antiviral effector (OAS2), IFN-type I-induced (EIF2AK2), adenosine deaminase (ADAR), vitamin A receptor family (RARB), the innate antiviral factor (TRIM5 and TRIM22), mucin-like protocadherin (MUPCDH) and other genes. These identified variants reside in genes spanning a broad collection of different classes of immune-related genes. Since genetic polymorphisms may function jointly to determine the outcome of vaccine-induced antibody response, it is reasonable to suggest that the observed antibody level effects in our study may be an outcome of combinations of SNP-defined alleles and immune response pathways, with the cumulative sum of variants across these genes influencing levels of rubella-specific antibodies. Better understanding the immunogenetic impact of multiple gene family pathways critical to development of humoral immune responses following rubella vaccination may provide insight into the factors that influence rubella immunity.
Strengths and Limitations
This manuscript outlines the first analysis of its kind, an assessment of the associations between a large collection of SNPs from immunologically relevant genes and rubella antibody levels. This analysis made it possible to examine all SNPs we genotyped on a well-characterized cohort of study participants and assess the likelihood that the univariate SNP results were significant on their own. This examination also made it possible to study a variety of types of associations between candidate SNPs and rubella antibody levels, both epistatic and otherwise. When complicated analytical approaches were employed, we used resampling methodologies to obtain statistical significance levels that were not over-trained to the data . This study does have two important limitations, however. First, as with all genetic association studies there is a need for replication, and a replication cohort is currently unavailable to us. Second, there was limited power to detect significant epistatic associations; for two SNPs with MAFs of 0.15 there was 80% power to detect interactions if the effect of carrying one copy of each of the minor alleles was associated with a 1.8-fold departure from additivity, while the detectable effect for a main effect for these SNPs was for a 1.25-fold difference. An additional limitation of this study is that our analyses were restricted to a collection of 726 tagSNPs identified from 84 candidate genes. There are likely to be other genetic variants that contribute to differences in rubella antibody levels that we have not assessed.
The implications of this work are two-fold. First, it presents a novel methodological approach that can provide additional insights into candidate genetic associations when analyzing similar data in other studies. When searching for associations between genes that are part of a complex system and outcomes that are potentially influenced by genetic variation in this system, it is important to consider more than just simple single-SNP assessments. Second, it is clear that the "dominant allele" model of genetic regulation of complex system processes such as immune response is both uncommon and too limiting. At least for those viral vaccines we have studied (measles, mumps, rubella, vaccinia and influenza), variation in immune responses are clearly the result of small contributions from many genes and sets of genes acting in concert. These data provide motivation to move beyond simple univariate associations to the types of analyses illustrated in this report in order to identify significant multigenic associations otherwise unobservable by simple univariate models. The value of approaches such as this include the ability to more clearly identify and model the genetic determinants of immune response to viral vaccines. The findings from these models can then be used to inform the development of next-generation vaccines.
The characteristics of the 738 healthy children and young adults (age 11 to 19 years) who participated in this study have previously been described . Study subjects were immunized with two age-appropriate doses of live measles-mumps-rubella-II (MMR-II) vaccine containing the Wistar RA 27/3-strain of rubella virus. The study participants were residents of Olmsted County, Minnesota, a community where no case of rubella infection had been reported during their lifetimes. The majority of the study population was white (91%), with 47% being female, with a median age at enrollment of 15 years (Table 1). The median age at the first and second immunization were 15 months and 11 years, respectively, and the median time between last rubella immunization and sample draw was 5.8 years. While 738 children and young adults were enrolled in the study, after exclusion for genotyping quality a total of 714 subjects were retained for analyses in this study. The Mayo Clinic Institutional Review Board granted approval for the study, and written informed consent (parental permission and assent for minors) was obtained.
Humoral Immunity Assays
Rubella-specific IgG antibodies after two doses of the rubella vaccine were detected in serum by a whole rubella virus-specific chemiluminescent immunoassay (Beckman Coulter Access, Fullerton, CA) according to the manufacturer's instructions. The limit of detection for this assay was 0.5 IU/ml and the coefficient of variation in our laboratory was 6%.
We selected a total of 768 tagSNPs from 84 candidate genes encoding cytokines and cytokine receptors. The details of SNP selection have described previously . Briefly, we generated a list of SNPs within, and 10 kb upstream and downstream of, the targeted candidate genes using the Hapmap Phase II (http://www.hapmap.org), Seattle SNPs (http://pga.mbt.washington.edu/), and NIEHS SNPs (http://egp.gs.washington.edu/) as source databases. We included SNPs that had validation data, successful predictive genotyping scores for Illumina GoldenGate assays, and reported minor allele frequencies (MAF) ≥0.05 in the list of possible SNPs. We used the linkage disequilibrium (LD) based ldSelect algorithm  to identify tagSNPs from the initial list with pairwise LD threshold of r2 <0.90 for Caucasians. Using these criteria, we selected 768 potential SNPs in our candidate genes of interest (see Additional file 1). We used the nomenclature described by den Dunnen and Antonarakis for all genotype variants .
Our genotyping methods were previously described in detail . Briefly, genomic DNA samples (n = 738, 250 ng each) obtained from frozen blood clots using the Puregene extraction kit (Gentra Systems Inc., Minneapolis, MN) were genotyped for 768 candidate SNPs using a custom designed 768-plex Illumina GoldenGate™ assay (Illumina Inc., San Diego, CA), following the manufacturer's instructions . All the SNPs selected for the custom Illumina panel had design scores >0.4. A Coriel Trio DNA (mother: NA11875, father: NA10859, daughter: NA10858) and two other genomic DNA controls were used as standards to review and refine clustering. These controls were genotyped on each plate, which allowed us to assess genotyping concordance of replicate subjects.
We used PCR-based TaqMan assays (Applied Biosystems, Foster City, CA) as the secondary platform to genotype SNPs that failed genotyping on the Illumina platform. All assays were performed according to the manufacturer's instructions, and the results were analyzed on the ABI Prism 7900 using Sequence Detection Software (Applied Biosystems).
Illumina 10% GenCall scores >0.4 and call rates >90% were used as thresholds for the initial laboratory quality control. The data from genotype calls made by BeadStudio 2 software were transferred to SAS for further analysis, where call rates for each subject, and for each SNP, were estimated. Participants' genotypes were used to estimate allele frequencies for each SNP, and departures from Hardy-Weinberg equilibrium (HWE) were assessed using a Pearson goodness-of-fit test or, for SNPs with a minor allele frequency of less than 5%, a Fisher exact test .
The purpose of the efforts reported here was to determine the multigenic contributions of SNPs from the candidate genes involved with immunity that were selected for this study. Prior to performing statistical comparisons, the characteristics of study participants were categorized and examined in relationship to the measured rubella-specific IgG antibody levels (represented as IU/ml). The median, and the 25th and 75th percentiles, of the antibody levels were summarized overall, and for the categories of the descriptive variables. Comparisons of the antibody levels across the different categorizations of the descriptive variables were achieved via analysis of variance (ANOVA), with rubella antibody levels being analyzed on the logarithmic scale to correct for data skewness. All analyses were adjusted for the following set of covariates potentially associated with the measured rubella-specific antibody titers: age at enrollment, race, gender, age at first rubella vaccination, age at second rubella vaccination, and cohort status.
While the focus of this effort was on the multigenic contributions of these SNPs to variation in rubella-specific antibody levels, a test of association was obtained for each genotyped SNP that met quality control criteria in our study subjects to examine the ordinal genetic association between each of the SNPs of interest while adjusting for the descriptive variables using linear regression. A quantile plot of the overall distribution of single-SNP p-values was extracted and summarized for all SNPs.
The primary interest was to identify the degree to which combinations of SNPs might provide novel insight into the potential control of rubella-specific antibody levels. Therefore, a series of multigenic analyses were carried out. We first assessed the degree to which individual SNPs provided additive information into the levels of rubella-specific antibody titers. We applied two approaches to identify these multi-allelic associations. The first was a single omnibus test of association where we identified the positively associated allele of each SNP and tallied the number of positively associated alleles of the candidate SNPs within each individual and assessed the association between this count variable and rubella antibody levels while controlling for the descriptive variables. The second approach again forced the inclusion of descriptive covariates, but relied on a stepwise model selection paradigm for identification of SNPs to be included. In this stepwise process, we added or removed a single ordinal SNP variable at a time until only SNPs that contributed to an association with rubella antibody levels remained in the regression model, using p-values for entrance or removal of 0.05. For each of these methods, we applied randomization tests to account for the fact that we identified the "positively associated allele" for each SNP within the same dataset in which we were performing the test of significance, and to account for the model selection approach we used to identify the model of candidate SNPs . These randomization procedures rely on repeated random re-assignment of observed antibody levels to model the situation where there is no association between SNPs and phenotype. These repeated randomization data sets are then used to perform the analyses that were carried out on the original data. The resulting p-values from these randomized data sets are then used to obtain an empirical estimate of the null distribution of p-values. The final randomization p-value reflects how extreme the observed p-value is, relative to this null empirical distribution.
In order to search for potential combinations of SNPs that were jointly associated with antibody levels, we applied the approach advocated by Marchini et al , in which we enumerated all possible combinations of the genotypes observed for each pair of candidate SNPs and tested for differences in the log-transformed rubella antibody levels among these categories using linear regression methods while adjusting for the descriptive variables. After performing this test for each of the 263,175 pairs of candidate SNPs, we produced a quantile plot to summarize the difference between the observed p-value distribution and p-value distribution that would have been expected if there were no pairs of SNPs that were significantly associated with antibody levels.
Additionally, we used recursive partitioning techniques to search for potential higher-order interactions among the candidate SNPs that might influence the observed levels of rubella-specific antibody levels. For a quantitative phenotype, these methods recursively search for optimal combinations of splits of explanatory variables such that the differences in the average levels of the phenotype among the identified groups are maximized [47, 48]. We utilized the recursive partitioning routines available in S-Plus (TIBCO Software Inc., Palo Alto, CA) to perform these assessments. All statistical tests were two-sided and, unless otherwise indicated, analyses were carried out using the SAS software system (SAS Institute, Inc., Cary, NC).
Conflict of Interest Statement
Dr. Poland is the chair of a safety evaluation committee for novel non-rubella vaccines undergoing clinical studies by Merck Research Laboratories.
VSP Dr. Pankratz is an Associate Professor of Biostatistics with training in statistical genetics. His interests lie in the study of genetic associations with a variety of phenotypes. He has worked with the Mayo Vaccine Research Group since 1998, studying the genetics of immune responses following vaccination.
RAV Mr. Vierkant is an Assistant Professor of Biostatistics with extensive experience in the design and analysis of genetic association studies. He has worked with the Mayo Vaccine Research Group since 1998, studying the genetics of immune responses following vaccination.
MMO Ms. O'Byrne is a Statistical Programmer Analyst in the Division of Biomedical Statistics and Informatics with extensive experience in analyzing data, and in preparing code to run a broad variety of statistical analyses, both novel and standard.
IGO Dr. Ovsyannikova is an Associate Professor of Medicine with a broad background in immunology and vaccinology. She has extensive experience in the design and conduct of genetics studies that focus on the contribution of candidate genes to differences in a variety of measures of post-vaccination immune response.
GAP Dr. Poland is a Professor of Medicine, Molecular Pharmacology & Experimental Therapeutics, Infectious Diseases. He has led the Mayo Clinic Vaccine Research Group since 1988 and helped to quantify the degree of genetic contribution to variation in a number of post-vaccination immune response measures. The efforts of the group he leads have focused in recent years on the study of genetic associations with a broad range of immune response measures raised by a variety of vaccines.
We thank the Mayo Clinic Vaccine Research Group and subjects who participated in our studies. We would like to thank nurses and study coordinators of the Mayo Clinic Vaccine Research Group for their efforts in recruiting subjects for this study. This work was supported by NIH grants AI 48793, AI 33144 and 1 UL1 RR024150-01 from the National Center for Research Resources (NCRR), a component of the National Institutes of Health, and the NIH Roadmap for Medical Research. Its contents are solely the responsibility of the authors and do not necessarily represent the official view of NCRR or NIH. While the results reported here were obtained from the application of a novel series of analyses to a large collection of SNPs, several of the candidate associations had been identified univariately, and had been published elsewhere [10, 11, 15, 16]. These associations are cited in the text of the manuscript. Several other components of this research were presented in part at the 49th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, California, 12-15 September 2009 (abstracts G1-891 and G1-1110).
- Best JM, Castillo-Solorzano C, Spika JS, Icenogle J, Glasser JW, Gay NJ, Andrus J, Arvin AM: Reducing the Global Burden of Congenital Rubella Syndrome: Report of the World Health Organization Steering Committee on Research Related to Measles and Rubella Vaccines and Vaccination, June 2004. J Infect Dis. 2005, 192: 1890-7. 10.1086/497607.View ArticlePubMedGoogle Scholar
- Novo A, Huebschen JM, Muller CP, Tesanovic M, Bojanic J: Ongoing rubella outbreak in Bosnia and Herzegovina, March-July 2009--preliminary report. Euro Surveill. 2009, 14: 19343-PubMedGoogle Scholar
- Tan P-L, Jacobson RM, Poland GA, Jacobsen SJ, Pankratz VS: Twin studies of immunogenicity - determining the genetic contribution to vaccine failure. Vaccine. 2001, 19: 2434-2439. 10.1016/S0264-410X(00)00468-0.View ArticlePubMedGoogle Scholar
- Kato S, Kimura M, Takakura I, Sakakibara T, Inouye H, Tsuji K: Possible associations between HLA antigens and the immune responsiveness to attenuated rubella vaccine. Tissue Antigens. 1978, 11: 475-478. 10.1111/j.1399-0039.1978.tb01287.x.View ArticlePubMedGoogle Scholar
- Ovsyannikova IG, Jacobson RM, Ryan JE, Dhiman N, Vierkant RA, Poland GA: Relationship between HLA polymorphisms and gamma interferon and interleukin-10 cytokine production in healthy individuals after rubella vaccination. Clin Vaccine Immunol. 2007, 14: 115-122. 10.1128/CVI.00247-06.PubMed CentralView ArticlePubMedGoogle Scholar
- Ou D, Mitchell LA, Décarie D, Tingle AJ, Nepom GT: Promiscuous T-cell recognition of a rubella capsid protein epitope restricted by DRB1*0403 and DRB1*0901 molecules sharing an HLA DR supertype. Hum Immunol. 1998, 59: 149-157. 10.1016/S0198-8859(98)00006-8.View ArticlePubMedGoogle Scholar
- Ovsyannikova IG, Jacobson RM, Vierkant RA, Jacobsen SJ, Pankratz VS, Poland GA: The contribution of HLA class I antigens in immune status following two doses of rubella vaccination. Hum Immunol. 2004, 65: 1506-1515. 10.1016/j.humimm.2004.07.001.View ArticlePubMedGoogle Scholar
- Ovsyannikova IG, Jacobson RM, Vierkant RA, Jacobsen SJ, Pankratz VS, Poland GA: Human leukocyte antigen class II alleles and rubella-specific humoral and cell-mediated immunity following measles-mumps-rubella-II vaccination. J Infect Dis. 2005, 191: 515-519. 10.1086/427558.View ArticlePubMedGoogle Scholar
- Ovsyannikova IG, Jacobson RM, Vierkant RA, O'Byrne MM, Poland GA: Replication of rubella vaccine population genetic studies: validation of HLA genotype and humoral response associations. Vaccine. 2009, 27: 6926-31. 10.1016/j.vaccine.2009.08.109.PubMed CentralView ArticlePubMedGoogle Scholar
- Ovsyannikova IG, Dhiman N, Haralambieva IH, Vierkant RA, O'Byrne MM, Jacobson RM, Poland GA: Rubella vaccine-induced cellular immunity: Evidence of associations with polymorphisms in the Toll-like, Vitamin A and D receptors and innate immune response genes. Hum Genet. 2010, 127: 207-221. 10.1007/s00439-009-0763-1.PubMed CentralView ArticlePubMedGoogle Scholar
- Ovsyannikova IG, Haralambieva IH, Dhiman N, O'Byrne MM, Pankratz VS, Jacobson RM, Poland GA: Polymorphisms in the vitamin a receptor and innate immunity genes influence the antibody response to rubella vaccination. J Infect Dis. 2010, 201: 207-213. 10.1086/649588.PubMed CentralView ArticlePubMedGoogle Scholar
- Dhiman N, Ovsyannikova I, Vierkant R, Pankratz VS, Jacobson R, Poland GA: Associations between cytokine/cytokine receptor SNPs and humoral immunity to measles, mumps and rubella in a Somali population. Tissue Antigens. 2008, 72: 211-220. 10.1111/j.1399-0039.2008.01097.x.PubMed CentralView ArticlePubMedGoogle Scholar
- Ovsyannikova IG, Pankratz VS, Vierkant RA, Jacobson RM, Poland GA: Human Leukocyte Antigen Haplotypes in the Genetic Control of Immune Response to Measles-Mumps-Rubella Vaccine. J Infect Dis. 2006, 193: 655-663. 10.1086/500144.View ArticlePubMedGoogle Scholar
- Kennedy RB, Ovsyannikova IG, Vierkant RA, Jacobson RM, Poland GA: Effect of human leukocyte antigen homozygosity on rubella vaccine-induced humoral and cell-mediated immune responses. Hum Immunol. 2010, 71: 128-135. 10.1016/j.humimm.2009.11.002.PubMed CentralView ArticlePubMedGoogle Scholar
- Haralambieva IH, Dhiman N, Ovsyannikova IG, Vierkant RA, Pankratz VS, Jacobson RM, Poland GA: 2'-5'-Oligoadenylate synthetase single-nucleotide polymorphisms and haplotypes are associated with variations in immune responses to rubella vaccine. Human Immunology. 2010, 71: 383-391. 10.1016/j.humimm.2010.01.004.PubMed CentralView ArticlePubMedGoogle Scholar
- Dhiman N, Haralambieva IH, Kennedy RB, Vierkant RA, O'Byrne MM, Ovsyannikova IG, Jacobson RM, Poland GA: SNP/haplotype associations in cytokine and cytokine receptor genes and immunity to rubella vaccination. Immunogenetics. 2010, 62: 197-210. 10.1007/s00251-010-0423-6.PubMed CentralView ArticlePubMedGoogle Scholar
- Dhiman N, Ovsyannikova IG, Cunningham JM, Vierkant RA, Kennedy RB, Pankratz VS, Poland GA, Jacobson RM: Associations between measles vaccine immunity and single nucleotide polymorphisms in cytokine and cytokine receptor genes. J Infect Dis. 2007, 195: 21-29. 10.1086/510596.View ArticlePubMedGoogle Scholar
- Ovsyannikova IG, Jacobson RM, Dhiman N, Vierkant RA, Pankratz VS, Poland GA: Human leukocyte antigen and cytokine receptor gene polymorphisms associated with heterogeneous immune responses to mumps viral vaccine. Pediatrics. 2008, 121: e1091-e1099. 10.1542/peds.2007-1575.PubMed CentralView ArticlePubMedGoogle Scholar
- van Deventer SJH: Cytokine and cytokine receptor polymorphisms in infectious disease. Intensive Care Med. 2000, 26 (Suppl 1): S98-S102. 10.1007/s001340051125.View ArticlePubMedGoogle Scholar
- Haukim N, Bidwell JL, Smith AJP, Keen LJ, Gallagher G, Kimberly R, Huizinga T, McDermott MF, Oksenberg J, McNicholl J, et al.,: Cytokine gene polymorphism in human disease: on-line databases, Supplement 2. Genes and Immunity. 2002, 3: 313-330. 10.1038/sj.gene.6363881.View ArticlePubMedGoogle Scholar
- Hill AV: The immunogenetics of human infectious diseases. Annu Rev Immunol. 1998, 16: 593-617. 10.1146/annurev.immunol.16.1.593.View ArticlePubMedGoogle Scholar
- Liu PT, Stenger S, Li H, Wenzel L, Tan BH, Krutzik SR, Ochoa MT, Schauber J, Wu K, Meinken C, et al.,: Toll-like receptor triggering of a vitamin D-mediated human antimicrobial response. Science. 2006, 311: 1770-1773. 10.1126/science.1123933.View ArticlePubMedGoogle Scholar
- Bochud PY, Bochud M, Telentim A, Calandra T: Innate immunogenetics: a tool for exploring new frontiers of host defence. Lancet Infect Dis. 2007, 7: 531-542. 10.1016/S1473-3099(07)70185-8.View ArticlePubMedGoogle Scholar
- MacKichan ML: Toll bridge to immunity. Immune molecules hold promise for vaccine adjuvant discovery. IAVI Rep. 2005, 9: 1-5.PubMedGoogle Scholar
- Mora JR, Iwata M, von Andrian UH: Vitamin effects on the immune system: vitamins A and D take centre stage. Nat Rev Immunol. 2008, 8: 685-698. 10.1038/nri2378.PubMed CentralView ArticlePubMedGoogle Scholar
- Poland GA, Ovsyannikova IG, Jacobson RM, Smith DI: Heterogeneity in vaccine immune response: The role of immunogenetics and the emerging field of vaccinomics. Clin Pharmacol Ther. 2007, 82: 653-664. 10.1038/sj.clpt.6100415.View ArticlePubMedGoogle Scholar
- Trowsdale J, Parham P: Mini-review: defense strategies and immunity-related genes. Eur J Immunol. 2004, 34: 7-17. 10.1002/eji.200324693.View ArticlePubMedGoogle Scholar
- Petranyi GG: The complexity of immune and alloimmune response. Transpl Immunol. 2002, 10: 91-100. 10.1016/S0966-3274(02)00054-0.View ArticlePubMedGoogle Scholar
- Motsinger AA, Haas DW, Hulgan T, Ritchie MD: Human genomic association studies: A primer for the infectious diseases specialist. J Infect Dis. 2007, 195: 1737-1744. 10.1086/518247.View ArticlePubMedGoogle Scholar
- Gibson G: Epistasis and pleiotropy as natural properties of transcriptional regulation. Theor Popul Biol. 1996, 49: 58-89. 10.1006/tpbi.1996.0003.View ArticlePubMedGoogle Scholar
- Thornton-Wells TA, Moore JH, Haines JL: Genetics, statistics and human disease: analytical retooling for complexity. Trends Genet. 2004, 20: 640-647. 10.1016/j.tig.2004.09.007.View ArticlePubMedGoogle Scholar
- Hoh J, Wille A, Ott J: Trimming, weighting, and grouping SNPs in human case-control association studies. Genome Research. 2001, 11: 2115-2119. 10.1101/gr.204001.PubMed CentralView ArticlePubMedGoogle Scholar
- Nelson MR, Kardia SL, Ferrell RE, Sing CF: A combinatorial partitioning method to identify multilocus genotypic partitions that predict quantitative trait variation. Genome Research. 2001, 11: 458-470. 10.1101/gr.172901.PubMed CentralView ArticlePubMedGoogle Scholar
- Ritchie MD, Hahn LW, Roodi N, Bailey LR, Dupont WD, Parl FF, Moore JH: Multifactor dimensionality reduction reveals high-order interactions among estrogen-metabolism genes in sporadic breast cancer. Am J Hum Genet. 2001, 69: 138-147. 10.1086/321276.PubMed CentralView ArticlePubMedGoogle Scholar
- Millstein J, Conti DV, Gilliland FD, Gauderman WJ: A testing framework for identifying susceptibility genes in the presence of epistasis. Am J Hum Genet. 2006, 78: 15-27. 10.1086/498850.PubMed CentralView ArticlePubMedGoogle Scholar
- Motsinger AA, Dudek SM, Hahn LW, Ritchie MD: Comparison of neural network optimization approaches for studies of human genetics. Lect Notes Comput Sci. 2006, 3907: 103-104. full_text.View ArticleGoogle Scholar
- Schaid DJ, McDonnell SK, Hebbring SJ, Cunningham JM, Thibodeau SN: Nonparametric tests of association of multiple genes with human disease. Am J Hum Genet. 2005, 76: 780-793. 10.1086/429838.PubMed CentralView ArticlePubMedGoogle Scholar
- Marchini J, Donnelly P, Cardon LR: Genome-wide strategies for detecting multiple loci that influence complex diseases. Nature Genet. 2005, 37: 413-417. 10.1038/ng1537.View ArticlePubMedGoogle Scholar
- Manly BFJ: Randomization, Bootstrap and Monte Carlo Methods in Biology. 1997, Boca Raton, FL: Chapman & Hall/CRC, 2Google Scholar
- Ovsyannikova IG, Ryan JE, Vierkant RA, O'Byrne MM, Pankratz VS, Jacobson RM, Poland GA: Influence of host genetic variation on rubella-specific T cell cytokine responses following rubella vaccination. Vaccine. 2009, 27: 3359-3366. 10.1016/j.vaccine.2009.01.079.PubMed CentralView ArticlePubMedGoogle Scholar
- The International HapMap Project. City.
- SeattleSNPs Variation Discovery Resource. City.
- National Institute of Environmental Health Sciences Environmental Genome Project: NIEHS SNPs.
- Carlson CS, Eberle MA, Rieder MJ, Yi Q, Kruglyak L, Nickerson DA: Selecting a maximally informative set of single-nucleotide polymorphisms for association analyses using linkage disequilibrium. Am J Hum Genet. 2004, 74: 106-120. 10.1086/381000.PubMed CentralView ArticlePubMedGoogle Scholar
- den Dunnen JT, Antonarakis SE: Nomenclature for the description of human sequence variations. Hum Genet. 2001, 109: 121-124. 10.1007/s004390100505.View ArticlePubMedGoogle Scholar
- Weir B: Genetic data Analysis II: methods for discrete population genetic data. 1996, Sunderland, MA: Sinauer Associates, Inc, 98-99.Google Scholar
- Breiman L, Friedman JH, Olshen RA, Stone CJ: Classification and Regression Trees. 1983, Belmont, CA: WadsworthGoogle Scholar
- Clark LA, Pregibon D: Tree-based models. Statistical Models in S. Edited by: Chambers JM, Hastie TJ. 1992, Pacific Grove, CA: Wadsworth and Brooks/ColeGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.