Statistical Methods for Differential Expressions of Genes Detected in Multiple-Condition Experiment of Microarray

Most studies of differential gene-expressions have been conducted between two given conditions. The two-condition experimental (TCE) approach is simple in that all genes detected display a common differential expression pattern responsive to a common two-condition difference. Therefore, the genes that are differentially expressed under the other conditions other than the given two conditions are undetectable with the TCE approach. In order to address the problem, we propose a new approach called multiple-condition experiment (MCE) without replication and develop corresponding statistical methods including inference of pairs of conditions for genes, new t-statistics, and a generalized multiple-testing method for any multiple-testing procedure via a control parameter C. We applied these statistical methods to analyze our real MCE data from breast cancer cell lines and found that 85 percent of gene-expression variations were caused by genotypic effects and genotype-ANAX1 overexpression interactions, which agrees well with our expected results. We also applied our methods to the adenoma dataset of Notterman et al. and identified 93 differentially expressed genes that could not be found in TCE. The MCE approach is a conceptual breakthrough in many aspects: (a) many conditions of interests can be conducted simultaneously; (b) study of association between differential expressions of genes and conditions becomes easy; (c) it can provide more precise information for molecular classification and diagnosis of tumors; (d) it can save lot of experimental resources and time for investigators.^

Analysis of residuals in contingency tables: another nail in the coffin of conditional approaches to significance testing.

Omnibus tests of significance in contingency tables use statistics of the chi-square type. When the null is rejected, residual analyses are conducted to identify cells in which observed frequencies differ significantly from expected frequencies. Residual analyses are thus conditioned on a significant omnibus test. Conditional approaches have been shown to substantially alter type I error rates in cases involving t tests conditional on the results of a test of equality of variances, or tests of regression coefficients conditional on the results of tests of heteroscedasticity. We show that residual analyses conditional on a significant omnibus test are also affected by this problem, yielding type I error rates that can be up to 6 times larger than nominal rates, depending on the size of the table and the form of the marginal distributions. We explored several unconditional approaches in search for a method that maintains the nominal type I error rate and found out that a bootstrap correction for multiple testing achieved this goal. The validity of this approach is documented for two-way contingency tables in the contexts of tests of independence, tests of homogeneity, and fitting psychometric functions. Computer code in MATLAB and R to conduct these analyses is provided as Supplementary Material.

Advanced Adhesion Strength Testing Methods of Thin Film Multilayers in Electronic Packaging Systems

With the continued miniaturization and increasing performance of electronic devices, new technical challenges have arisen. One such issue is delamination occurring at critical interfaces inside the device. This major reliability issue can occur during the manufacturing process or during normal use of the device. Proper evaluation of the adhesion strength of critical interfaces early in the product development cycle can help reduce reliability issues and time-to-market of the product. However, conventional adhesion strength testing is inherently limited in the face of package miniaturization, which brings about further technical challenges to quantify design integrity and reliability. Although there are many different interfaces in today's advanced electronic packages, they can be generalized into two main categories: 1) rigid to rigid connections with a thin flexible polymeric layer in between, or 2) a thin film membrane on a rigid structure. Knowing that every technique has its own advantages and disadvantages, multiple testing methods must be enhanced and developed to be able to accommodate all the interfaces encountered for emerging electronic packaging technologies. For evaluating the adhesion strength of high adhesion strength interfaces in thin multilayer structures a novel adhesion test configuration called “single cantilever adhesion test (SCAT)” is proposed and implemented for an epoxy molding compound (EMC) and photo solder resist (PSR) interface. The test method is then shown to be capable of comparing and selecting the stronger of two potential EMC/PSR material sets. Additionally, a theoretical approach for establishing the applicable testing domain for a four-point bending test method was presented. For evaluating polymeric films on rigid substrates, major testing challenges are encountered for reducing testing scatter and for factoring in the potentially degrading effect of environmental conditioning on the material properties of the film. An advanced blister test with predefined area test method was developed that considers an elasto-plastic analytical solution and implemented for a conformal coating used to prevent tin whisker growth. The advanced blister testing with predefined area test method was then extended by employing a numerical method for evaluating the adhesion strength when the polymer’s film properties are unknown.

Brain-derived neurotrophic factor (BDNF) gene : no major impact on antidepressant treatment response

The brain-derived neurotrophic factor (BDNF) has been suggested to play a pivotal role in the aetiology of affective disorders. In order to further clarify the impact of BDNF gene variation on major depression as well as antidepressant treatment response, association of three BDNF polymorphisms [rs7103411, Val66Met (rs6265) and rs7124442] with major depression and antidepressant treatment response was investigated in an overall sample of 268 German patients with major depression and 424 healthy controls. False discovery rate (FDR) was applied to control for multiple testing. Additionally, ten markers in BDNF were tested for association with citalopram outcome in the STAR*D sample. While BDNF was not associated with major depression as a categorical diagnosis, the BDNF rs7124442 TT genotype was significantly related to worse treatment outcome over 6 wk in major depression (p=0.01) particularly in anxious depression (p=0.003) in the German sample. However, BDNF rs7103411 and rs6265 similarly predicted worse treatment response over 6 wk in clinical subtypes of depression such as melancholic depression only (rs7103411: TT

Association study of the calcitonin gene-related polypeptide-alpha (CALCA) and the receptor activity modifying 1 (RAMP1) genes with migraine

Migraine is a common neurovascular brain disorder characterised by recurrent attacks of severe headache that may be accompanied by various neurological symptoms. Migraine is thought to result from activation of the trigeminovascular system followed by vasodilation of pain-producing intracranial blood vessels and activation of second-order sensory neurons in the trigeminal nucleus caudalis. Calcitonin gene-related peptide (CGRP) is a mediator of neurogenic inflammation and the most powerful vasodilating neuropeptide, and has been implicated in migraine pathophysiology. Consequently, genes involved in CGRP synthesis or CGRP receptor genes may play a role in migraine and/or increase susceptibility. This study investigates whether variants in the gene that encodes CGRP, calcitonin-related polypeptide alpha (CALCA) or in the gene that encodes a component of its receptor, receptor activity modifying protein 1 (RAMP1), are associated with migraine pathogenesis and susceptibility. The single nucleotide polymorphisms (SNPs) rs3781719 and rs145837941 in the CALCA gene, and rs3754701 and rs7590387 at the RAMP1 locus, were analysed in an Australian Caucasian population of migraineurs and matched controls. Although we find no significant association of any of the SNPs tested with migraine overall, we detected a nominally significant association (p = 0.031) of the RAMP1 rs3754701 variant in male migraine subjects, although this is non-significant after Bonferroni correction for multiple testing.

Association analysis of chromosome 1 migraine candidate genes

Migraine with aura (MA) is a subtype of typical migraine. Migraine with aura (MA) also encompasses a rare severe subtype Familial Hemiplegic Migraine (FHM) with several known genetic loci. The type 2 FHM (FHM-2) susceptibility locus maps to chromosome 1q23 and mutations in the ATP1A2 gene at this site have recently been implicated. We have previously provided evidence of linkage of typical migraine (predominantly MA) to microsatellite markers on chromosome 1, in the 1q31 and 1q23 regions. In this study, we have undertaken a large genomic investigation involving candidate genes that lie within the chromosome 1q23 and 1q31 regions using an association analysis approach. Methods We have genotyped a large population of case-controls (243 unrelated Caucasian migraineurs versus 243 controls) examining a set of 5 single nucleotide polymorphisms (SNPs) and the Fas Ligand dinucleotide repeat marker, located within the chromosome 1q23 and 1q31 regions. Results Several genes have been studied including membrane protein (ATP 1 subtype A4 and FasL), cytoplasmic glycoprotein (CASQ 1) genes and potassium (KCN J9 and KCN J10) and calcium (CACNA1E) channel genes in 243 migraineurs (including 85% MA and 15% of migraine without aura (MO)) and 243 matched controls. After correction for multiple testing, chi-square results showed non-significant P values (P > 0.008) across all SNPs (and a CA repeat) tested in these different genes, however results with the KCN J10 marker gave interesting results (P = 0.02) that may be worth exploring further in other populations. Conclusion These results do not show a significant role for the tested candidate gene variants and also do not support the hypothesis that a common chromosome 1 defective gene influences both FHM and the more common forms of migraine.

The GSTM1 null genotype confers an increased risk for Solar Keratosis Development in an Australian Caucasian Population

Solar keratoses affect approximately 50% of Australian Caucasians aged over 40 y. Solar keratoses can undergo malignant transformation into squamous cell carcinoma followed by possible metastasis and are risk factors for basal cell carcinoma, melanoma, and squamous cell carcinoma. The glutathione-S-transferase genes play a part in detoxification of carcinogens and mutagens, including some produced by ultraviolet radiation. This study examined the role of glutathione-S-transferase M1, T1, P1, and Z1 gene polymorphisms in susceptibility to solar keratoses development. Using DNA samples from volunteers involved in the Nambour Skin Cancer Prevention Trial, allele and genotype frequencies were determined using polymerase chain reaction and restriction enzyme digestion. No significant differences were detected in glutathione-S-transferase P1 and glutathione-S-transferase Z1 allele or genotype frequencies; however, a significant association between glutathione-S-transferase M1 genotypes and solar keratoses development was detected (p=0.003) with null individuals having an approximate 2-fold increase in risk for solar keratoses development (odds ratio: 2.1; confidence interval: 1.3-3.5) and a significantly higher increase in risk in conjunction with high outdoor exposure (odds ratio: 3.4; confidence interval: 1.9-6.3). Also, a difference in glutathione-S-transferase T1 genotype frequencies was detected (p=0.039), although considering that multiple testing was undertaken, this was found not to be significant. Fair skin and inability to tan were found to be highly significant risk factors for solar keratoses development with odds ratios of 18.5 (confidence interval: 5.7-59.9) and 7.4 (confidence interval: 2.6-21.0), respectively. Overall, glutathione-S-transferase M1 conferred a significant increase in risk of solar keratoses development, particularly in the presence of high outdoor exposure and synergistically with known phenotypic risk factors of fair skin and inability to tan.

Environmental, personal, and genetic determinants of response to vitamin D supplementation in older adults

CONTEXT AND OBJECTIVE: Suboptimal vitamin D status can be corrected by vitamin D supplementation, but individual responses to supplementation vary. We aimed to examine genetic and nongenetic determinants of change in serum 25-hydroxyvitamin D (25(OH)D) after supplementation. DESIGN AND PARTICIPANTS: We used data from a pilot randomized controlled trial in which 644 adults aged 60 to 84 years were randomly assigned to monthly doses of placebo, 30 000 IU, or 60 000 IU vitamin D3 for 12 months. Baseline characteristics were obtained from a self-administered questionnaire. Eighty-eight single-nucleotide polymorphisms (SNPs) in 41 candidate genes were genotyped using Sequenom MassArray technology. Serum 25(OH)D levels before and after the intervention were measured using the Diasorin Liaison platform immunoassay. We used linear regression models to examine associations between genetic and nongenetic factors and change in serum 25(OH)D levels. RESULTS: Supplement dose and baseline 25(OH)D level explained 24% of the variability in response to supplementation. Body mass index, self-reported health status, and ambient UV radiation made a small additional contribution. SNPs in CYP2R1, IRF4, MC1R, CYP27B1, VDR, TYRP1, MCM6, and HERC2 were associated with change in 25(OH)D level, although only CYP2R1 was significant after adjustment for multiple testing. Models including SNPs explained a similar proportion of variability in response to supplementation as models that included personal and environmental factors. CONCLUSION: Stepwise regression analyses suggest that genetic variability may be associated with response to supplementation, perhaps suggesting that some people might need higher doses to reach optimal 25(OH)D levels or that there is variability in the physiologically normal level of 25(OH)D.

Dysbindin (DTNBP1) variants are associated with hallucinations in schizophrenia

BACKGROUND: Dystrobrevin binding protein 1 (DTNBP1) is a schizophrenia susceptibility gene involved with neurotransmission regulation (especially dopamine and glutamate) and neurodevelopment. The gene is known to be associated with cognitive deficit phenotypes within schizophrenia. In our previous studies, DTNBP1 was found associated not only with schizophrenia but with other psychiatric disorders including psychotic depression, post-traumatic stress disorder, nicotine dependence and opiate dependence. These findings suggest that DNTBP1 may be involved in pathways that lead to multiple psychiatric phenotypes. In this study, we explored the association between DTNBP1 SNPs (single nucleotide polymorphisms) and multiple psychiatric phenotypes included in the Diagnostic Interview of Psychosis (DIP). METHODS: Five DTNBP1 SNPs, rs17470454, rs1997679, rs4236167, rs9370822 and rs9370823, were genotyped in 235 schizophrenia subjects screened for various phenotypes in the domains of depression, mania, hallucinations, delusions, subjective thought disorder, behaviour and affect, and speech disorder. SNP-phenotype association was determined with ANOVA under general, dominant/recessive and over-dominance models. RESULTS: Post hoc tests determined that SNP rs1997679 was associated with visual hallucination; SNP rs4236167 was associated with general auditory hallucination as well as specific features including non-verbal, abusive and third-person form auditory hallucinations; and SNP rs9370822 was associated with visual and olfactory hallucinations. SNPs that survived correction for multiple testing were rs4236167 for third-person and abusive form auditory hallucinations; and rs9370822 for olfactory hallucinations. CONCLUSION: These data suggest that DTNBP1 is likely to play a role in development of auditory related, visual and olfactory hallucinations which is consistent with evidence of DTNBP1 activity in the auditory processing regions, in visual processing and in the regulation of glutamate and dopamine activity

No genetic support for a contribution of prostaglandins to the aetiology of androgenetic alopecia

MADAM, Androgenetic alopecia (AGA) is a common age-dependent trait, characterized by a progressive loss of hair from the scalp. The hair loss may commence during puberty and up to 80% of white men experience some degree of AGA during their lifetime.1 Research has established that two essential aetiological factors for AGA are a genetic predisposition and the presence of androgens (male sex hormones).1,2 A recent meta-analysis of genome-wide association studies (GWAS) has increased the number of identified loci associated with this trait at the molecular level to a total of eight.3 However, despite these successes, a large fraction of the genetic contribution remains to be identified. One way to identify further genetic loci is to combine the resource of GWAS datasets with knowledge about specific biological factors likely to be involved in the development of disease. The focused evaluation of a limited number of candidate genes in GWAS datasets avoids the necessity for extensive correction for multiple testing, which typically limits the power for detecting genetic loci at a genome-wide level.4 Because the presence of genetic association suggests that candidate genes are likely to operate early in the causative chain of events leading to the phenotype, this approach may also function to favour biological pathways for their importance in the development of AGA.

A genome-wide association study of sleep habits and insomnia

Several aspects of sleep behavior such as timing, duration and quality have been demonstrated to be heritable. To identify common variants that influence sleep traits in the population, we conducted a genome-wide association study of six sleep phenotypes assessed by questionnaire in a sample of 2,323 individuals from the Australian Twin Registry. Genotyping was performed on the Illumina 317, 370, and 610K arrays and the SNPs in common between platforms were used to impute non-genotyped SNPs. We tested for association with more than 2,000,000 common polymorphisms across the genome. While no SNPs reached the genome-wide significance threshold, we identified a number of associations in plausible candidate genes. Most notably, a group of SNPs in the third intron of the CACNA1C gene ranked as most significant in the analysis of sleep latency (P = 1.3 x 10(-)(6)). We attempted to replicate this association in an independent sample from the Chronogen Consortium (n = 2,034), but found no evidence of association (P = 0.73). We have identified several other suggestive associations that await replication in an independent sample. We did not replicate the results from previous genome-wide analyses of self-reported sleep phenotypes after correction for multiple testing.

Association mapping

Association mapping seeks to identify marker alleles present at significantly different frequencies in cases carrying a particular disease or trait compared with controls. Genome-wide association studies are increasingly replacing candidate gene-based association studies for complex diseases, where a number of loci are likely to contribute to disease risk and the effect size of each particular risk allele is typically modest or low. Good study design is essential to the success of an association study, and factors such as the heritability of the disease under investigation, the choice of controls, statistical power, multiple testing and whether the association can be replicated need to be considered before beginning. Likewise, thorough quality control of the genotype data needs to be undertaken prior to running any association analyses. Finally, it should be kept in mind that a significant genetic association is not proof positive that a particular genetic locus causes a disease, but rather an important first step in discovering the genetic variants underlying a complex disease.

Educational attainment: A genome wide association study in 9538 Australians

BACKGROUND Correlations between Educational Attainment (EA) and measures of cognitive performance are as high as 0.8. This makes EA an attractive alternative phenotype for studies wishing to map genes affecting cognition due to the ease of collecting EA data compared to other cognitive phenotypes such as IQ. METHODOLOGY In an Australian family sample of 9538 individuals we performed a genome-wide association scan (GWAS) using the imputed genotypes of approximately 2.4 million single nucleotide polymorphisms (SNP) for a 6-point scale measure of EA. Top hits were checked for replication in an independent sample of 968 individuals. A gene-based test of association was then applied to the GWAS results. Additionally we performed prediction analyses using the GWAS results from our discovery sample to assess the percentage of EA and full scale IQ variance explained by the predicted scores. RESULTS The best SNP fell short of having a genome-wide significant p-value (p = 9.77x10(-7)). In our independent replication sample six SNPs among the top 50 hits pruned for linkage disequilibrium (r(2)<0.8) had a p-value<0.05 but only one of these SNPs survived correction for multiple testing--rs7106258 (p = 9.7*10(-4)) located in an intergenic region of chromosome 11q14.1. The gene based test results were non-significant and our prediction analyses show that the predicted scores explained little variance in EA in our replication sample. CONCLUSION While we have identified a polymorphism chromosome 11q14.1 associated with EA, further replication is warranted. Overall, the absence of genome-wide significant p-values in our large discovery sample confirmed the high polygenic architecture of EA. Only the assembly of large samples or meta-analytic efforts will be able to assess the implication of common DNA polymorphisms in the etiology of EA.

Variation in BMPR1B, TGFRB1 and BMPR2 and control of dizygotic twinning

Genes in the TGF9 signaling pathway play important roles in the regulation of ovarian follicle growth and ovulation rate. Mutations in three genes in this pathway, growth differentiation factor 9 (GDF9), bone morphogenetic protein 15 (BMP15) and the bone morphogenetic protein receptor B 1 (BMPRB1), influence dizygotic (DZ) twinning rates in sheep. To date, only variants in GDF9 and BMP15, but not their receptors transforming growth factor ss receptor 1 (TGFBR1), bone morphogenetic protein receptor 2 (BMPR2) and BMPR1B, have been investigated with respect to their roles in human DZ twinning. We screened for rare and novel variants in TGFBR1, BMPR2 and BMPR1B in mothers of dizygotic twins (MODZT) from twin-dense families, and assessed association between genotyped and imputed variants and DZ twinning in another large sample of MODZT. Three novel variants were found: a deep intronic variant in BMPR2, and one intronic and one non-synonymous exonic variant in BMPRB1 which would result in the replacement of glutamine by glutamic acid at amino acid position 294 (p.Gln294Glu). None of these variants were predicted to have major impacts on gene function. However, the p.Gln294Glu variant changes the same amino acid as a sheep BMPR1B functional variant and may have functional consequences. Six BMPR1B variants were marginally associated with DZ twinning in the larger case-control sample, but these were no longer significant once multiple testing was taken into account. Our results suggest that variation in the TGF9 signaling pathway type II receptors has limited effects on DZ twinning rates in humans.

Novel Insights into the Molecular Genetic Background of Colorectal Tumors

Many of the genes predisposing to highly penetrant colorectal cancer (CRC) syndromes, including hereditary non-polyposis colorectal cancer (MLH1, MSH2, MSH6, PMS2), familial adenomatous polyposis (APC), Peutz-Jeghers syndrome (LKB1), juvenile polyposis (SMAD4, BMPR1A), MYH-associated polyposis (MYH), and Cowden syndrome (PTEN) have already been discovered. Identification of these genes has allowed a more precise classification of the hereditary CRC syndromes and provided a means for predictive genetic testing and surveillance. Some of the genes are also involved in sporadic cancer forms, and therefore the investigation of the rare CRC syndromes has been a breakthrough for general cancer research. Despite the accumulating knowledge on hereditary cancer syndromes, a significant number of familial CRCs remain molecularly unexplained after genetic testing, reflecting the possibility of other predisposing genes or existence of novel syndromes. Moreover, genetic variants conferring low-penetrance risk are still largely unknown. In this study, we examined the role of some new high- and low-penetrance alleles on CRC predisposition. We identified disease causing MYH mutations in a subset (9%) of patients with APC and AXIN2 mutation negative adenomatous polyposis. Due to differences in the pattern of inheritance and clinical manifestation, screening for mutations in MYH is beneficial in view of genetic counselling and surveillance. A novel functionally deficient MYH founder mutation A459D was identified in the Finnish population, and this finding had immediate clinical implications for genetic counselling of at risk families. Many patients with hamartomatous polyposis remain without molecular diagnosis due to atypical phenotypes. We therefore sought to classify 49 patients with unexplained hamartomatous or hyperplastic/mixed polyposis by extensive molecular analyses of PTEN, LKB1, BMPR1A, SMAD4, ENG, BRAF, MYH, and BHD along with revision of polyp histology. Mutations were identified in 11/49 (22%) of the patients. In 6 cases the molecular diagnosis was re-classified guiding surveillance and decisions for prophylactic surgery. Re-evaluation of polyp histology with subsequent more accurate selection of candidate gene analyses is beneficial and can be recommended for patients with unexplained polyposis. Furthermore, germline mutations in ENG underlying juvenile polyposis were described for the first time, characterizing a possible novel genetically defined form of hereditary CRC. Association analyses on two putative low-penetrance alleles, NOD2 3020insC and MDM2 SNP309 were performed in a population-based series of 1042 Finnish CRC patients and in cancer-free controls. In contrast to previous results, NOD2 3020insC did not associate with CRC or age at disease onset in the Finnish population. These data suggest that NOD2 3020insC alone might not be sufficient for CRC predisposition. MDM2 SNP309 was as common in the CRC cohort as in the healthy controls. Interesting trends, however, were observed, which after correction for multiple testing did not reach statistical significance. SNP309 was more common in female CRC patients and a trend towards an earlier age at disease onset was observed in women with SNP309. Subsequent studies have supported this observation and SNP309 could affect gender- or hormone-related tumorigenesis. Finally, a large-scale unbiased effort was designed to characterize the complete mutatome of CRC with microsatellite instability (MSI). Using an approach combining expression microarray and genome database searches, we were able to identify putative MSI target genes. Further characterization of one of the genes suggested that it might play a role also in microsatellite stable CRC and Peutz-Jeghers syndrome pathogenesis.