Association mapping of complex diseases in linked regions: estimation of genetic effects and feasibility of testing rare variants

Principles in genetic risk assessment

Risk assessment constitutes an essential component of genetic counseling and testing, and the genetic risk should be estimated as accurately as possible for individual and family decision making. All relevant information retrieved from population studies and pedigree and genetic testing enhances the accuracy of the assessment of an individual's genetic risk. This review will focus on the following general aspects implicated in risk assessment: the increasing genetic information regarding disease; complex traits versus Mendelian disorders; and the influence of the environment and disease susceptibility. The influence of these factors on risk assessment will be discussed.

Familial risks for common diseases: etiologic clues and guidance to gene identification

Familial clustering of a disease is a direct indicator of a possible heritable cause, provided that environmental sharing can be excluded. If the familial clustering is lacking, the likelihood of a heritable influence is also small. In the era of genome scans, the consideration of data on heritability should be important in the assessment of the likely success of the genome scan. The availability of a Multigeneration Register in Sweden provides a reliable access to families throughout the last century. This Register has been extensively used to study a number of different diseases through linkage to the Hospital Discharge Register. In the present article we review the obtained and some unpublished results for nine main disease classes. For each of these, familial risks are given for four disease subtypes. As measures of familial clustering we use risks between siblings, twins and spouses. Disease correlation between spouses suggests environmental sharing and a higher correlation between siblings and particularly twins shows heritable effects. We will also comment on the established susceptibility genes and the risks conferred by them. The data suggest high heritabilities for chronic obstructive pulmonary disease, asthma, noninfective enteritis and colitis, cerebral palsy and endocrine and metabolic diseases. Among the performed first-generation genome scans on various diseases, the success appears to be related to the a priori heritability estimates. To our knowledge this is a first attempt to summarize familial risks for a large number of diseases using data from a single population on which reasonable uniform diagnostic criteria have been applied.

Two-dimensional linkage analyses of rheumatoid arthritis

Rheumatoid arthritis (RA) is a multifactorial disease with complex genetic etiology, about which little is known. Here, we apply a two-stage procedure in which a quick first-stage analysis was used to narrow down targets for a more thorough and detailed testing for gene x gene interaction. Potentially interesting regions were first identified by testing for major gene effects using non-parametric linkage methods. To select regions of interest, we first tested for linkage to three different RA-related traits one at a time: RA affection status and the quantitative phenotypes rheumatoid factor IgM and anti-cyclic citrullinated peptide levels. These linkage analyses identified regions on chromosomes 3, 5, 6, 8, 16, 18, 19, and 20. We subsequently analyzed the selected regions in a pairwise manner to detect gene x gene interactions influencing RA using a recently developed two-dimensional linkage method. We found evidence of interacting loci on chromosomes 5, 6, and 18.

The balance between heritable and environmental aetiology of human disease

The Human Genome Project and the ensuing International HapMap Project were largely motivated by human health issues. But the distance from a DNA sequence variation to a novel disease gene is considerable; for complex diseases, closing this gap hinges on the premise that they arise mainly from heritable causes. Using cancer as an example of complex disease, we examine the scientific evidence for the hypothesis that human diseases result from interactions between genetic variants and the environment.

Design and analysis of genetic association studies to finely map a locus identified by linkage analysis: sample size and power calculations

Association (e.g. case-control) studies are often used to finely map loci identified by linkage analysis. We investigated the influence of various parameters on power and sample size requirements for such a study. Calculations were performed for various values of a high-risk functional allele (fA), frequency of a marker allele associated with the high risk allele (f1), degree of linkage disquilibrium between functional and marker alleles (D') and trait heritability attributable to the functional locus (h2). The calculations show that if cases and controls are selected from equal but opposite extreme quantiles of a quantitative trait, the primary determinants of power are h2 and the specific quantiles selected. For a dichotomous trait, power also depends on population prevalence. Power is optimal if functional alleles are studied (fA= f1 and D'= 1.0) and can decrease substantially as D' diverges from 1.0 or as f(1) diverges from fA. These analyses suggest that association studies to finely map loci are most powerful if potential functional polymorphisms are identified a priori or if markers are typed to maximize haplotypic diversity. In the absence of such information, expected minimum power at a given location for a given sample size can be calculated by specifying a range of potential frequencies for fA (e.g. 0.1-0.9) and determining power for all markers within the region with specification of the expected D' between the markers and the functional locus. This method is illustrated for a fine-mapping project with 662 single nucleotide polymorphisms in 24 Mb. Regions differed by marker density and allele frequencies. Thus, in some, power was near its theoretical maximum and little additional information is expected from additional markers, while in others, additional markers appear to be necessary. These methods may be useful in the analysis and interpretation of fine-mapping studies.

Relationships between familial risks of cancer and the effects of heritable genes and their SNP variants

Familial risks for cancer can be used in many ways in guiding gene identification efforts and, more broadly, in understanding cancer etiology. Gene identification efforts may be properly designed and targeted if the familial risks are well characterized and the mode of inheritance is identified. Single nucleotide polymorphisms (SNPs) are extensively used in case-control studies of practically all cancer types. They are used for the identification of inherited cancer susceptibility genes and those that may interact with environmental factors. However, being genetic markers, they are applicable only on heritable conditions, which is often a neglected fact. Based on the data in the nationwide Swedish Family-Cancer Database, we review familial risks for all main cancers and discuss the evidence for a heritable component in cancer. The available evidence, including differences in cancer incidence between regions and temporal changes within regions, indicates that cancer is mainly an environmental disease, with a minor heritable etiology. The large environmental component will hamper the success of SNP-based genetic association studies. Empirical familial risks should be used to evaluate the feasibility of such studies. We develop figures for the assessment of genetic parameters based on familial risks. Such data are helpful in the estimation of the expected genetic effects in cancer. Overall, we consider the likelihood of a successful application of SNPs in gene-environment studies small, unless established environmental risk factors are tested on proven candidate genes.

Recent developments in genomewide association scans: a workshop summary and review

With the imminent availability of ultra-high-volume genotyping platforms (on the order of 100,000-1,000,000 genotypes per sample) at a manageable cost, there is growing interest in the possibility of conducting genomewide association studies for a variety of diseases but, so far, little consensus on methods to design and analyze them. In April 2005, an international group of >100 investigators convened at the University of Southern California over the course of 2 days to compare notes on planned or ongoing studies and to debate alternative technologies, study designs, and statistical methods. This report summarizes these discussions in the context of the relevant literature. A broad consensus emerged that the time was now ripe for launching such studies, and several common themes were identified--most notably the considerable efficiency gains of multistage sampling design, specifically those made by testing only a portion of the subjects with a high-density genomewide technology, followed by testing additional subjects and/or additional SNPs at regions identified by this initial scan.

No evidence for association of the TATA-box binding protein glutamine repeat sequence or the flanking chromosome 6q27 region with type 1 diabetes

Susceptibility to the autoimmune disease type 1 diabetes has been linked to human chromosome 6q27 and, moreover, recently associated with one of the genes in the region, TATA box-binding protein (TBP). Using a much larger sample of T1D families than those studied by others, and by extensive re-sequencing of nine other genes in the proximity, in which we identified 279 polymorphisms, 83 of which were genotyped in up to 725 T1D multiplex and simplex families, we obtained no evidence for association of the TBP CAG/CAA (glutamine) microsatellite repeat sequence with disease, or for nine other genes, PDCD2, PSMB1, KIAA1838, DLL1, dJ894D12.4, FLJ25454, FLJ13162, FLJ11152, PHF10 and CCR6. This study also provides an exon-based tag single nucleotide polymorphism map for these 10 genes that can be used for analysis of other diseases.

Genome-wide association studies: theoretical and practical concerns

To fully understand the allelic variation that underlies common diseases, complete genome sequencing for many individuals with and without disease is required. This is still not technically feasible. However, recently it has become possible to carry out partial surveys of the genome by genotyping large numbers of common SNPs in genome-wide association studies. Here, we outline the main factors - including models of the allelic architecture of common diseases, sample size, map density and sample-collection biases - that need to be taken into account in order to optimize the cost efficiency of identifying genuine disease-susceptibility loci.

The allelic spectra of common diseases may resemble the allelic spectrum of the full genome

Identification of the genes responsible for common human diseases promises to be one of the most significant advances in medical knowledge and treatment. To date, the numerous attempts to identify the genes responsible for complex and multi-factorial common diseases have met with only a handful of successes. The key to calculating the optimal effort and ideal approach to successful identifications lies with understanding the likely allelic spectrum of the target disease. The allelic spectrum describes the number of disease loci and the frequency of each disease allele. It has been implicitly assumed that disease spectra are biased towards either commonness or rareness relative to the allelic spectrum of the overall human genome. We present a hypothesis that the allelic spectra of common diseases are generally similar to the spectrum that characterizes the entire genome. This hypothesis is supported by the fact that only a few loci have major significance to familial disease risks and that there may be many disease loci which each make a minor contribution to a disease. Additionally, although relatively few alleles of the human genome have been examined for disease involvement, current estimates of the number of disease genes are very high. Because selection will have been operating only weakly and for a relatively short time on most of the alleles associated with complex diseases, spectra that are characteristic of near-neutral selection may well apply. We thus propose that the hitherto neglected hypothesis that puts the likely allelic spectra of common diseases in the middle ground between the prevailing hypotheses of spectral skew towards rareness or commonness is the most likely. By using this hypothesis as the null, research resources may be optimally allocated and greater success in identifying disease genes may be achieved.

Genetic epidemiology of cancer: From families to heritable genes

A reliable determination of familial risks for cancer is important for clinical counseling, prevention and understanding cancer etiology. Family-based gene identification efforts may be targeted if the risks are well characterized and the mode of inheritance is identified. Medically verified data on familial risks have not been available for all types of cancer but they have become available through the use of the nationwide Swedish Family-Cancer Database, which includes all Swedes born in 1932 and later with their parents, totaling over 10 million individuals. Over 150 publications have emanated from this source. The familial risks of cancer have been characterized for all main cancers and the contribution of environmental and heritable effects to the familial aggregation has been assessed. Furthermore, the mode of inheritance has been deduced by comparing risks from parental and sibling probands. Examples are shown on familial clustering of cancers, for which heritable susceptibility genes are yet unknown, such as squamous cell carcinoma of the skin, intestinal carcinoids, thyroid papillary tumors, brain astrocytomas and pituitary adenomas. Some common cancers, such as lung and kidney cancers, appear to show an early-onset recessive component because familial risks among siblings are much higher than those in families where parents are probands. Many of the cancer sites showing high familial risks lack guidelines for clinical counseling or action level. In conclusion, we recommend that any future gene identification efforts, either using linkage or association designs, devise their strategies based on data from family studies. Clinical genetic counseling would benefit from reviewing established familial risks on all main types of cancer.

New methods for finding disease-susceptibility genes: impact and potential

Improved techniques for defining disease-gene location and evaluating the biological candidacy of regional transcripts will hasten disease-gene discovery.

Improved techniques for defining disease-gene location and evaluating the biological candidacy of regional transcripts will hasten disease-gene discovery.

Linkage and association mapping of the LRP5 locus on chromosome 11q13 in type 1 diabetes

Linkage of chromosome 11q13 to type 1 diabetes (T1D) was first reported from genome scans (Davies et al. 1994; Hashimoto et al. 1994) resulting in P <2.2 x 10(-5) (Luo et al. 1996) and designated IDDM4 ( insulin dependent diabetes mellitus 4). Association mapping under the linkage peak using 12 polymorphic microsatellite markers suggested some evidence of association with a two-marker haplotype, D11S1917*03-H0570POLYA*02, which was under-transmitted to affected siblings and over-transmitted to unaffected siblings ( P=1.5 x 10(-6)) (Nakagawa et al. 1998). Others have reported evidence for T1D association of the microsatellite marker D11S987, which is approximately 100 kb proximal to D11S1917 (Eckenrode et al. 2000). We have sequenced a 400-kb interval surrounding these loci and identified four genes, including the low-density lipoprotein receptor related protein (LRP5) gene, which has been considered as a functional candidate gene for T1D (Hey et al. 1998; Twells et al. 2001). Consequently, we have developed a comprehensive SNP map of the LRP5 gene region, and identified 95 SNPs encompassing 269 kb of genomic DNA, characterised the LD in the region and haplotypes (Twells et al. 2003). Here, we present our refined linkage curve of the IDDM4 region, comprising 32 microsatellite markers and 12 SNPs, providing a peak MLS=2.58, P=5 x 10(-4), at LRP5 g.17646G>T. The disease association data, largely focused in the LRP5 region with 1,106 T1D families, provided no further evidence for disease association at LRP5 or at D11S987. A second dataset, comprising 1,569 families from Finland, failed to replicate our previous findings at LRP5. The continued search for the variants of the putative IDDM4 locus will greatly benefit from the future development of a haplotype map of the genome.