Bayesian methodology for genetics of complex diseases

Genetic research of complex diseases is a challenging, but exciting, area of research. The early development of the research was limited, however, until the completion of the Human Genome and HapMap projects, along with the reduction in the cost of genotyping, which paves the way for understanding the genetic composition of complex diseases. In this thesis, we focus on the statistical methods for two aspects of genetic research: phenotype definition for diseases with complex etiology and methods for identifying potentially associated Single Nucleotide Polymorphisms (SNPs) and SNP-SNP interactions. With regard to phenotype definition for diseases with complex etiology, we firstly investigated the effects of different statistical phenotyping approaches on the subsequent analysis. In light of the findings, and the difficulties in validating the estimated phenotype, we proposed two different methods for reconciling phenotypes of different models using Bayesian model averaging as a coherent mechanism for accounting for model uncertainty. In the second part of the thesis, the focus is turned to the methods for identifying associated SNPs and SNP interactions. We review the use of Bayesian logistic regression with variable selection for SNP identification and extended the model for detecting the interaction effects for population based case-control studies. In this part of study, we also develop a machine learning algorithm to cope with the large scale data analysis, namely modified Logic Regression with Genetic Program (MLR-GEP), which is then compared with the Bayesian model, Random Forests and other variants of logic regression.

DRD2/ANKK1 Taq1A (rs 1800497 C>T) genotypes are associated with susceptibility to second generation antipsychotic-induced akathisia

Rationale Although the advent of atypical, second-generation antipsychotics (SGAs) has resulted in reduced likelihood of akathisia, this adverse effect remains a problem. Extrapyramidal adverse effects are associated with increased drug occupancy of dopamine 2 receptors (DRD2). The A1 allele of the DRD2/ANKK1,rs1800497, is associated with decreased striatal DRD2 density. Objectives The aim of this study was to identify whether the A1(T) allele of the DRD2/ANKK1 was associated with akathisia (measured with the Barnes Akathisia Rating Scale) in a clinical sample of 234 patients treated with antipsychotics. Results Definite akathisia (a score≥ 2 for the global clinical assessment of akathisia) was significantly less common in subjects prescribed SGAs (16.8 %) than those prescribed FGAs (47.6%), p<0.0001. Overall, 24.1% of A1+ (A1A2/A1A1) patients treated with SGAs had akathisia compared to 10.8% of A1- (A2A2) patients. A1+ (A1A2/A1A1) patients administered SGAs also had higher global clinical assessment of akathisia scores than A1- subjects (p=0.01). SGAs maintained their advantage over FGAs regarding akathisia even in A1+ patients treated with SGAs. Conclusions These results strongly suggest that A1+ variants of the DRD2/ANKK1 Taq1A allele confer risk for akathisia in patients treated with SGAs and may explain inconsistencies across prior studies comparing FGAs and SGAs.

Genetic polymorphisms in the human tissue kallikrein (KLK) locus and their implication in various malignant and non-malignant diseases

The Kallikrein (KLK) gene locus encodes a family of serine proteases and is the largest contiguous cluster of protease-encoding genes attributed an evolutionary age of 330 million years. The KLK locus has been implicated as a high susceptibility risk loci in numerous cancer studies through the last decade. The KLK3 gene already has established clinical relevance as a biomarker in prostate cancer prognosis through its encoded protein, prostate-specific antigen. Data mined through genome-wide association studies (GWAS) and next-generation sequencing point to many important candidate single nucleotide polymorphisms (SNPs) in KLK3 and other KLK genes. SNPs in the KLK locus have been found to be associated with several diseases including cancer, hypertension, cardiovascular disease and atopic dermatitis. Moreover, introducing a model incorporating SNPs to improve the efficiency of prostate-specific antigen in detecting malignant states of prostate cancer has been recently suggested. Establishing the functional relevance of these newly-discovered SNPs, and their interactions with each other, through in silico investigations followed by experimental validation, can accelerate the discovery of diagnostic and prognostic biomarkers. In this review, we discuss the various genetic association studies on the KLK loci identified either through candidate gene association studies or at the GWAS and post-GWAS front to aid researchers in streamlining their search for the most significant, relevant and therapeutically promising candidate KLK gene and/or SNP for future investigations.

Mapping quantitative trait loci in a wild population using linkage and linkage disequilibrium analyses

Historical information can be used, in addition to pedigree, traits and genotypes, to map quantitative trait locus (QTL) in general populations via maximum likelihood estimation of variance components. This analysis is known as linkage disequilibrium (LD) and linkage mapping, because it exploits both linkage in families and LD at the population level. The search for QTL in the wild population of Soay sheep on St. Kilda is a proof of principle. We analysed the data from a previous study and confirmed some of the QTLs reported. The most striking result was the confirmation of a QTL affecting birth weight that had been reported using association tests but not when using linkage-based analyses. Copyright © Cambridge University Press 2010.

Candidate gene analysis for quantitative traits using the transmission disequilibrium test : the example of the melanocortin 4-receptor in pigs

Population-wide associations between loci due to linkage disequilibrium can be used to map quantitative trait loci (QTL) with high resolution. However, spurious associations between markers and QTL can also arise as a consequence of population stratification. Statistical methods that cannot differentiate between loci associations due to linkage disequilibria from those caused in other ways can render false-positive results. The transmission-disequilibrium test (TDT) is a robust test for detecting QTL. The TDT exploits within-family associations that are not affected by population stratification. However, some TDTs are formulated in a rigid-form, with reduced potential applications. In this study we generalize TDT using mixed linear models to allow greater statistical flexibility. Allelic effects are estimated with two independent parameters: one exploiting the robust within-family information and the other the potentially biased between-family information. A significant difference between these two parameters can be used as evidence for spurious association. This methodology was then used to test the effects of the fourth melanocortin receptor (MC4R) on production traits in the pig. The new analyses supported the previously reported results; i.e., the studied polymorphism is either causal of in very strong linkage disequilibrium with the causal mutation, and provided no evidence for spurious association.

Transcriptomics approaches to identify genes and markers for production traits in giant freshwater prawn Macrobrachium rosenbergii

This study used next generation sequencing technologies to investigate growth in a cultured crustacean. The objective was to identify and characterise specific gene loci that contribute important phenotypic variation to growth as well as to develop a large set of SNP markers in candidate genes for assessing correlations between specific mutations and individual growth performance. The genomic dataset generated provides a fundamental resource for application in future crustacean stock improvement programs. Ultimately, the data can be applied to development of culture lines with improved growth performance.

A genome-wide analysis of 'Bounty' descendants implicates several novel variants in migraine susceptibility

Migraine is a common neurological disease with a complex genetic aetiology. The disease affects ~12% of the Caucasian population and females are three times more likely than males to be diagnosed. In an effort to identify loci involved in migraine susceptibility, we performed a pedigree-based genome-wide association study of the isolated population of Norfolk Island, which has a high prevalence of migraine. This unique population originates from a small number of British and Polynesian founders who are descendents of the Bounty mutiny and forms a very large multigenerational pedigree (Bellis et al.; Human Genetics, 124(5):543-5542, 2008). These population genetic features may facilitate disease gene mapping strategies (Peltonen et al.; Nat Rev Genet, 1(3):182-90, 2000. In this study, we identified a high heritability of migraine in the Norfolk Island population (h (2) = 0.53, P = 0.016). We performed a pedigree-based GWAS and utilised a statistical and pathological prioritisation approach to implicate a number of variants in migraine. An SNP located in the zinc finger protein 555 (ZNF555) gene (rs4807347) showed evidence of statistical association in our Norfolk Island pedigree (P = 9.6 × 10(-6)) as well as replication in a large independent and unrelated cohort with >500 migraineurs. In addition, we utilised a biological prioritisation to implicate four SNPs, in within the ADARB2 gene, two SNPs within the GRM7 gene and a single SNP in close proximity to a HTR7 gene. Association of SNPs within these neurotransmitter-related genes suggests a disrupted serotoninergic system that is perhaps specific to the Norfolk Island pedigree, but that might provide clues to understanding migraine more generally.

Comparing genotyping algorithms for Illumina's Infinium whole-genome SNP BeadChips

Background Illumina's Infinium SNP BeadChips are extensively used in both small and large-scale genetic studies. A fundamental step in any analysis is the processing of raw allele A and allele B intensities from each SNP into genotype calls (AA, AB, BB). Various algorithms which make use of different statistical models are available for this task. We compare four methods (GenCall, Illuminus, GenoSNP and CRLMM) on data where the true genotypes are known in advance and data from a recently published genome-wide association study. Results In general, differences in accuracy are relatively small between the methods evaluated, although CRLMM and GenoSNP were found to consistently outperform GenCall. The performance of Illuminus is heavily dependent on sample size, with lower no call rates and improved accuracy as the number of samples available increases. For X chromosome SNPs, methods with sex-dependent models (Illuminus, CRLMM) perform better than methods which ignore gender information (GenCall, GenoSNP). We observe that CRLMM and GenoSNP are more accurate at calling SNPs with low minor allele frequency than GenCall or Illuminus. The sample quality metrics from each of the four methods were found to have a high level of agreement at flagging samples with unusual signal characteristics. Conclusions CRLMM, GenoSNP and GenCall can be applied with confidence in studies of any size, as their performance was shown to be invariant to the number of samples available. Illuminus on the other hand requires a larger number of samples to achieve comparable levels of accuracy and its use in smaller studies (50 or fewer individuals) is not recommended.

G-protein β3 subunit gene (GNB3) variant in causation of essential hypertension

Essential hypertensives display enhanced signal transduction through pertussis toxin-sensitive G proteins. The T allele of a C825T variant in exon 10 of the G protein β3 subunit gene (GNB3) induces formation of a splice variant (Gβ3-s) with enhanced activity. The T allele of GNB3 was shown recently to be associated with hypertension in unselected German patients (frequency=0.31 versus 0.25 in control). To confirm and extend this finding in a different setting, we performed an association study in Australian white hypertensives. This involved an extensively examined cohort of 110 hypertensives, each of whom were the offspring of 2 hypertensive parents, and 189 normotensives whose parents were both normotensive beyond age 50 years. Genotyping was performed by polymerase chain reaction and digestion with BseDI, which either cut (C allele) or did not cut (T allele) the 268-bp polymerase chain reaction product. T allele frequency in the hypertensive group was 0.43 compared with 0.25 in the normotensive group (χ2=22; P=0.00002; odds ratio=2.3; 95% CI=1.7 to 3.3). The T allele tracked with higher pretreatment blood pressure: diastolic=105±7, 109±16, and 128±28 mm Hg (mean±SD) for CC, CT, and 7T, respectively (P=0.001 by 1-way ANOVA). Blood pressures were higher in female hypertensives with a T allele (P=0.006 for systolic and 0.0003 for diastolic by ANOVA) than they were in male hypertensives. In conclusion, the present study of a group with strong family history supports a role for a genetically determined, physiologically active splice variant of the G protein β3 subunit gene in the causation of essential hypertension.

Common PPARγ variants C161T and Pro12Ala are not associated with inflammatory bowel disease in an Australian cohort

Background & Aims: Peroxisome proliferator-activated receptor (PPAR) γ is a transcription factor, highly expressed in colonic epithelial cells, adipose tissue and macrophages, with an important role in the regulation of inflammatory pathways. The common PPARγ variants C161T and Pro12Ala have recently been associated with Ulcerative Colitis (UC) and an extensive UC phenotype respectively, in a Chinese population. PPARγ Pro12Ala variant homozygotes appear to be protected from the development of Crohn's disease (CD) in European Caucasians. Methods: A case-control study was performed for both variants (CD n=575, UC n=306, Controls n=360) using a polymerase chain reaction (PCR)-restriction fragment length polymorphism analysis in an Australian IBD cohort. A transmission disequilibrium test was also performed using CD trios for the PPARγ C161T variant. Genotype-phenotype analyses were also undertaken. Results: There was no significant difference in genotype distribution data or allele frequency between CD and UC patients and controls. There was no difference in allele transmission for the C161T variant. No significant relationship between the variants and disease location was observed. Conclusions: We were unable to replicate in a Caucasian cohort the recent association between PPARγ C161T and UC or between PPARγ Pro12Ala and an extensive UC phenotype in a Chinese population. There are significant ethnic differences in genetic susceptibility to IBD and its phenotypic expression.

An investigation of the 5-HT2C receptor gene as a migraine candidate gene

Migraine is a common complex disorder, currently classified into two main subtypes, migraine with aura (MA) and migraine without aura (MO). The strong preponderance of females to males suggests an X-linked genetic component. Recent studies have identified an X chromosomal susceptibility region (Xq24-q28) in two typical migraine pedigrees. This region harbours a potential candidate gene for the disorder, the serotonin receptor 2C (5-HT2C) gene. This study involved a linkage and association approach to investigate two single nucleotide variants in the 5-HT2C gene. In addition, exonic coding regions of the 5-HT2C gene were also sequenced for mutations in X-linked migraine pedigrees. Results of this study did not detect any linkage or association, and no disease causing mutations were identified. Hence, results for this study do not support a significant role of the 5-HT 2C gene in migraine predisposition. © 2003 Wiley-Liss, Inc.

Analysis of Sdhd and Mmp12 in an affected solar keratosis and control cohort

The incidence of Squamous Cell Carcinoma (SCG) is growing in certain populations to the extent that it is now the most common skin lesion in young men and women in high ultraviolet exposure regions such as Queensland. In terms of incidence up to 40% of the Australian population over 40 years of age is thought to possess the precancerous Solar Keratosis (SK) lesion and with a small, but significant, chance of progression into SCC, understanding the genetic events that play a role in this process is essential. The major aims of this study were to analyse whole blood derived samples for DNA aberrations in genes associated with tumour development and cellular maintenance, with the ultimate aim of identifying genes associated with non-melanoma skin cancer development. More specifically the first aim of this project was to analyse the SDHD and MMP12 genes via Dual-Labelled Probe Real-Time PCR for copy number aberrations in an affected Solar Keratosis and control cohort. It was found that 12 samples had identifiable copy-number aberrations in either the SDHD or MMP12 gene (this means that a genetic section of either of these two genes is aberrantly amplified or deleted), with five of the samples exhibiting aberrations in both genes. The significance of this study is the contribution to the knowledge of the genetic pathways that are malformed in the progression and development of the pre-cancerous skin lesion Solar Keratosis. © 2008 Springer Science+Business Media, LLC.

Independent, marked associations of alleles of the insulin receptor and dipeptidyl carboxypeptidase-I genes with essential hypertension

1. There is evidence to suggest that essential hypertension is a polygenic disorder and that it arises from yet-to-be-identified predisposing variants of certain genes that influence blood pressure. The cloning of various hormone, enzyme, adrenoceptor and hormone receptor genes whose products are involved in blood pressure control and the identification of polymorphisms of these has permitted us to test their genetic association with hypertension. 2. Cross-sectional analyses of a number of candidate gene markers were performed in hypertensive and normotensive subjects who were selected on the basis of both parents being either hypertensive or normotensive, respectively, and the difference in total alleles on all chromosomes for each polymorphism between the hypertensive and normotensive groups was test by χ analysis with one degree of freedom. 3. A marked association was observed between hypertension and insertion alleles of polymorphisms of the insulin receptor gene (INSR) (P<0.0040) and the dipeptidyl carboxypeptidase-1 (angiotensin I-converting enzyme; kininase II) gene (DCP1) (P<0.0018). No association with hypertension was evident, however, for polymorphisms of the growth hormone, low-density lipoprotein receptor, renal kallikrein, α2- and β1-adrenoreceptor, atrial natriuretic factor and insulin genes. 4. All but one of the hypertensive subjects had at least one of the hypertension-associated alleles, and although subjects homozygous for both were three times more frequent in the hypertensive group, examination of the nine possible genotypes suggested that the INSR and DCP1 alleles are independent markers for hypertension. 5. The present results suggest that genetic variant(s) in close linkage disequilibrium with polymorphisms at INSR and DCP1 may be involved in part in the aetiology of essential hypertension.

Prognostic value of TP53, KRAS and EGFR mutations in nonsmall cell lung cancer : the EUELC cohort

Nonsmall cell lung cancer samples from the European Early Lung Cancer biobank were analysed to assess the prognostic significance of mutations in the TP53, KRAS and EGFR genes. The series included 11 never-smokers, 86 former smokers, 152 current smokers and one patient without informed smoking status. There were 110 squamous cell carcinomas (SCCs), 133 adenocarcinomas (ADCs) and seven large cell carcinomas or mixed histologies. Expression of p53 was analysed by immunohistochemistry. DNA was extracted from frozen tumour tissues. TP53 mutations were detected in 48.8% of cases and were more frequent among SCCs than ADCs (p<0.0001). TP53 mutation status was not associated with prognosis. G to T transversions, known to be associated with smoking, were marginally more common among patients who developed a second primary lung cancer or recurrence/metastasis (progressive disease). EGFR mutations were almost exclusively found in never-smoking females (p=0.0067). KRAS mutations were detected in 18.5% of cases, mainly ADC (p<0.0001), and showed a tendency toward association with progressive disease status. These results suggest that mutations are good markers of different aetiologies and histopathological forms of lung cancers but have little prognostic value, with the exception of KRAS mutation, which may have a prognostic value in ADC. Copyright©ERS 2012.