A Classifier-based approach to identify genetic similarities between diseases

Circulating interleukins and risk of colorectal cancer: a Mendelian randomization study

BACKGROUND

Recent studies have suggested a potential causal association between Interleukins (ILs) and Colorectal Cancer (CRC), and thus, it is important to examine the causal relationship between them using a Mendelian randomization (MR) approach.

METHODS

The instrumental variables were extracted for IL-1ra, IL-6, IL-6ra, IL-8, IL-16, IL-18, IL-27 from genome-wide association studies of European ancestry. Summary statistics of CRC were also retrieved. An inverse variance-weighted MR approach was implemented as the primary method to compute overall effects from multiple instruments. Additional MR approaches and sensitivity and heterogeneity pleiotropy analyses were also conducted respectively.

RESULTS

Our analysis suggested a causal effect between an increase of IL-8 and a reduced risk of CRC (odds ratio 0.65; 95% confidence interval, 0.43-0.98; p = 0.041) and did not provide evidence for causal effects of IL-1ra, IL-6, IL-6ra, IL-16, IL-18, IL-27. Sensitivity analyses suggested the robustness of MR results and that they were unlikely to be affected by unbalanced pleiotropy or significant heterogeneity.

CONCLUSIONS

This study investigated the role of ILs in the development of CRC and we found a causal effect between an increase of IL-8 and a reduced risk of CRC but not found evidence for causal effects of IL-1ra, IL-6, IL-6ra, IL-16, IL-18, IL-27. Sensitivity analyses suggested the robustness of MR results and that they were unlikely to be affected by unbalanced pleiotropy or significant heterogeneity.

Evaluating individual genome similarity with a topic model

MOTIVATION

Evaluating genome similarity among individuals is an essential step in data analysis. Advanced sequencing technology detects more and rarer variants for massive individual genomes, thus enabling individual-level genome similarity evaluation. However, the current methodologies, such as the principal component analysis (PCA), lack the capability to fully leverage rare variants and are also difficult to interpret in terms of population genetics.

RESULTS

Here, we introduce a probabilistic topic model, latent Dirichlet allocation, to evaluate individual genome similarity. A total of 2535 individuals from the 1000 Genomes Project (KGP) were used to demonstrate our method. Various aspects of variant choice and model parameter selection were studied. We found that relatively rare (0.001<allele frequency < 0.175) and sparse (average interval > 20 000 bp) variants are more efficient for genome similarity evaluation. At least 100 000 such variants are necessary. In our results, the populations show significantly less mixed and more cohesive visualization than the PCA results. The global similarities among the KGP genomes are consistent with known geographical, historical and cultural factors.

AVAILABILITY AND IMPLEMENTATION

The source code and data access are available at: https://github.com/lrjuan/LDA_genome.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Comparative study for haplotype block partitioning methods - Evidence from chromosome 6 of the North American Rheumatoid Arthritis Consortium (NARAC) dataset

Haplotype-based methods compete with “one-SNP-at-a-time” approaches on being preferred for association studies. Chromosome 6 contains most of the known genetic biomarkers for rheumatoid arthritis (RA) disease. Therefore, chromosome 6 serves as a benchmark for the haplotype methods testing. The aim of this study is to test the North American Rheumatoid Arthritis Consortium (NARAC) dataset to find out if haplotype block methods or single-locus approaches alone can sufficiently provide the significant single nucleotide polymorphisms (SNPs) associated with RA. In addition, could we be satisfied with only one method of the haplotype block methods for partitioning chromosome 6 of the NARAC dataset? In the NARAC dataset, chromosome 6 comprises 35,574 SNPs for 2,062 individuals (868 cases, 1,194 controls). Individual SNP approach and three haplotype block methods were applied to the NARAC dataset to identify the RA biomarkers. We employed three haplotype partitioning methods which are confidence interval test (CIT), four gamete test (FGT), and solid spine of linkage disequilibrium (SSLD). P-values after stringent Bonferroni correction for multiple testing were measured to assess the strength of association between the genetic variants and RA susceptibility. Moreover, the block size (in base pairs (bp) and number of SNPs included), number of blocks, percentage of uncovered SNPs by the block method, percentage of significant blocks from the total number of blocks, number of significant haplotypes and SNPs were used to compare among the three haplotype block methods. Individual SNP, CIT, FGT, and SSLD methods detected 432, 1,086, 1,099, and 1,322 associated SNPs, respectively. Each method identified significant SNPs that were not detected by any other method (Individual SNP: 12, FGT: 37, CIT: 55, and SSLD: 189 SNPs). 916 SNPs were discovered by all the three haplotype block methods. 367 SNPs were discovered by the haplotype block methods and the individual SNP approach. The P-values of these 367 SNPs were lower than those of the SNPs uniquely detected by only one method. The 367 SNPs detected by all the methods represent promising candidates for RA susceptibility. They should be further investigated for the European population. A hybrid technique including the four methods should be applied to detect the significant SNPs associated with RA for chromosome 6 of the NARAC dataset. Moreover, SSLD method may be preferred for its favored benefits in case of selecting only one method.

Is Mania the Hypertension of the Mood? Discussion of A Hypothesis

Beyond both being biphasic/bidirectional disorders (hypo)mania and essential hypertension share a surprising number of similarities and an overlap between their genetics, biological background, underlying personality and temperamental factors, precipitating factors, comorbidity and response to treatment, indicating a possibly partially shared biological background. Based on theoretical knowledge, similarities related to characteristics, manifestation and course, and the results of pharmacological studies related to the effects and side effects of pharmacotherapies used in the treatment of these two distinct disorders, the authors outline a hypothesis discussing the similar origins of these two phenomena and thus mania being the hypertension of mood in memory of Athanasios Koukopoulos, one of the greatest researchers and theoreticists of mania of all time.

A comparative study of improvements Pre-filter methods bring on feature selection using microarray data

Background

Feature selection techniques have become an apparent need in biomarker discoveries with the development of microarray. However, the high dimensional nature of microarray made feature selection become time-consuming. To overcome such difficulties, filter data according to the background knowledge before applying feature selection techniques has become a hot topic in microarray analysis. Different methods may affect final results greatly, thus it is important to evaluate these pre-filter methods in a system way.

Methods

In this paper, we compared the performance of statistical-based, biological-based pre-filter methods and the combination of them on microRNA-mRNA parallel expression profiles using L1 logistic regression as feature selection techniques. Four types of data were built for both microRNA and mRNA expression profiles.

Results

Results showed that pre-filter methods could reduce the number of features greatly for both mRNA and microRNA expression datasets. The features selected after pre-filter procedures were shown to be significant in biological levels such as biology process and microRNA functions. Analyses of classification performance based on precision showed the pre-filter methods were necessary when the number of raw features was much bigger than that of samples. All the computing time was greatly shortened after pre-filter procedures.

Conclusions

With similar or better classification improvements, less but biological significant features, pre-filter-based feature selection should be taken into consideration if researchers need fast results when facing complex computing problems in bioinformatics.

Electronic supplementary material

The online version of this article (doi:10.1186/2047-2501-2-7) contains supplementary material, which is available to authorized users.

Principal component analysis characterizes shared pathogenetics from genome-wide association studies

Genome-wide association studies (GWASs) have recently revealed many genetic associations that are shared between different diseases. We propose a method, disPCA, for genome-wide characterization of shared and distinct risk factors between and within disease classes. It flips the conventional GWAS paradigm by analyzing the diseases themselves, across GWAS datasets, to explore their "shared pathogenetics". The method applies principal component analysis (PCA) to gene-level significance scores across all genes and across GWASs, thereby revealing shared pathogenetics between diseases in an unsupervised fashion. Importantly, it adjusts for potential sources of heterogeneity present between GWAS which can confound investigation of shared disease etiology. We applied disPCA to 31 GWASs, including autoimmune diseases, cancers, psychiatric disorders, and neurological disorders. The leading principal components separate these disease classes, as well as inflammatory bowel diseases from other autoimmune diseases. Generally, distinct diseases from the same class tend to be less separated, which is in line with their increased shared etiology. Enrichment analysis of genes contributing to leading principal components revealed pathways that are implicated in the immune system, while also pointing to pathways that have yet to be explored before in this context. Our results point to the potential of disPCA in going beyond epidemiological findings of the co-occurrence of distinct diseases, to highlighting novel genes and pathways that unsupervised learning suggest to be key players in the variability across diseases.

A Bayesian analysis of the chromosome architecture of human disorders by integrating reductionist data

In this paper, we present a Bayesian approach to estimate a chromosome and a disorder network from the Online Mendelian Inheritance in Man (OMIM) database. In contrast to other approaches, we obtain statistic rather than deterministic networks enabling a parametric control in the uncertainty of the underlying disorder-disease gene associations contained in the OMIM, on which the networks are based. From a structural investigation of the chromosome network, we identify three chromosome subgroups that reflect architectural differences in chromosome-disorder associations that are predictively exploitable for a functional analysis of diseases.

Genetic mapping of habitual substance use, obesity-related traits, responses to mental and physical stress, and heart rate and blood pressure measurements reveals shared genes that are overrepresented in the neural synapse

Links between substance use habits, obesity, stress and the related cardiovascular outcomes can be, in part, because of loci with pleiotropic effects. To investigate this hypothesis, we performed genome-wide mapping in 119 multigenerational families from a population in the Saguenay-Lac-St-Jean region with a known founder effect using 58 000 single-nucleotide polymorphisms and 437 microsatellite markers to identify genetic components of the following factors: habitual alcohol, tobacco and coffee use; response to mental and physical stress; obesity-related traits; and heart rate (HR) and blood pressure (BP) measures. Habitual alcohol and/or tobacco users had attenuated HR responses to mental stress compared with non-users, whereas hypertensive individuals had stronger HR and systolic BP responses to mental stress and a higher obesity index than normotensives. Genetic mappings uncovered numerous shared genes among substance use, stress response, obesity and hemodynamic traits, including CAMK4, CNTN4, DLG2, FHIT, GRID2, ITPR2, NOVA1 and PRKCE, forming network of interacting proteins, sharing synaptic function and display higher and patterned expression profiles in brain-related tissues; moreover, pathway analysis of shared genes pointed to long-term potentiation. Subgroup genetic mappings uncovered additional shared synaptic genes, including CAMK4, CNTN5 and DNM3 (hypertension-specific); CNTN4, DNM3, FHIT and ITPR1 (sex-specific), having protein interactions with genes driven from general analysis. In summary, consistent with the observed phenotypic correlations, we found substantial overlap among genomic determinants of these traits in synapse, which supports the notion that the neural synapse may be a shared interface behind substance use, stress, obesity, HR, BP as well as the observed sex- and hypertension-specific genetic differences.

Disease liability prediction from large scale genotyping data using classifiers with a reject option

Genome-wide association studies (GWA) try to identify the genetic polymorphisms associated with variation in phenotypes. However, the most significant genetic variants may have a small predictive power to forecast the future development of common diseases. We study the prediction of the risk of developing a disease given genome-wide genotypic data using classifiers with a reject option, which only make a prediction when they are sufficiently certain, but in doubtful situations may reject making a classification. To test the reliability of our proposal, we used the Wellcome Trust Case Control Consortium (WTCCC) data set, comprising 14,000 cases of seven common human diseases and 3,000 shared controls.

Evidence for somatic gene conversion and deletion in bipolar disorder, Crohn's disease, coronary artery disease, hypertension, rheumatoid arthritis, type-1 diabetes, and type-2 diabetes

BACKGROUND

During gene conversion, genetic information is transferred unidirectionally between highly homologous but non-allelic regions of DNA. While germ-line gene conversion has been implicated in the pathogenesis of some diseases, somatic gene conversion has remained technically difficult to investigate on a large scale.

METHODS

A novel analysis technique is proposed for detecting the signature of somatic gene conversion from SNP microarray data. The Wellcome Trust Case Control Consortium has gathered SNP microarray data for two control populations and cohorts for bipolar disorder (BD), cardiovascular disease (CAD), Crohn's disease (CD), hypertension (HT), rheumatoid arthritis (RA), type-1 diabetes (T1D) and type-2 diabetes (T2D). Using the new analysis technique, the seven disease cohorts are analyzed to identify cohort-specific SNPs at which conversion is predicted. The quality of the predictions is assessed by identifying known disease associations for genes in the homologous duplicons, and comparing the frequency of such associations with background rates.

RESULTS

Of 28 disease/locus pairs meeting stringent conditions, 22 show various degrees of disease association, compared with only 8 of 70 in a mock study designed to measure the background association rate (P < 10-9). Additional candidate genes are identified using less stringent filtering conditions. In some cases, somatic deletions appear likely. RA has a distinctive pattern of events relative to other diseases. Similarities in patterns are apparent between BD and HT.

CONCLUSIONS

The associations derived represent the first evidence that somatic gene conversion could be a significant causative factor in each of the seven diseases. The specific genes provide potential insights about disease mechanisms, and are strong candidates for further study.

Rethinking the genetic basis for comorbidity of schizophrenia and type 2 diabetes

The co-occurrence of schizophrenia (SCZ) and type 2 diabetes mellitus (T2D) has been well documented. This review article focuses on the hypothesis that the co-occurrence of SCZ and T2D may be, at least in part, driven by shared genetic factors. Previous genetic studies of T2D and SCZ evidence have disclosed a number of overlapped risk loci. However, the putative common genetic factors for SCZ and T2D remain inconclusive due to inconsistent findings. A systemic review of methods of identifying genetic loci contributing to the comorbidity link between SCZ and T2D is hence needed. In the current review article, we have discussed several different approaches to localizing the shared susceptibility genes for these two diseases. To begin with, one could start with probing the gene involved in both glucose and dopamine metabolisms. Additionally, hypothesis-free genome-wide association studies (GWAS) may provide more clues to the common genetic basis for these two diseases. Genetic similarities inferred from GWAS may shed some light on the genetic mechanism underlying the comorbidity link between SCZ and T2D. Meanwhile, endophenotypes (e.g., adiponectin level in T2D and working memory in SCZ) may serve as alternative phenotypes that are more directly influenced by genes than target diseases. Hence, endophenotypes of these diseases may be more tractable to identification. To summarize, novel approaches are needed to dissect the complex genetic basis of the comorbidity of SCZ and T2D.

Extreme evolutionary disparities seen in positive selection across seven complex diseases

Positive selection is known to occur when the environment that an organism inhabits is suddenly altered, as is the case across recent human history. Genome-wide association studies (GWASs) have successfully illuminated disease-associated variation. However, whether human evolution is heading towards or away from disease susceptibility in general remains an open question. The genetic-basis of common complex disease may partially be caused by positive selection events, which simultaneously increased fitness and susceptibility to disease. We analyze seven diseases studied by the Wellcome Trust Case Control Consortium to compare evidence for selection at every locus associated with disease. We take a large set of the most strongly associated SNPs in each GWA study in order to capture more hidden associations at the cost of introducing false positives into our analysis. We then search for signs of positive selection in this inclusive set of SNPs. There are striking differences between the seven studied diseases. We find alleles increasing susceptibility to Type 1 Diabetes (T1D), Rheumatoid Arthritis (RA), and Crohn's Disease (CD) underwent recent positive selection. There is more selection in alleles increasing, rather than decreasing, susceptibility to T1D. In the 80 SNPs most associated with T1D (p-value <7.01 x 10(-5)) showing strong signs of positive selection, 58 alleles associated with disease susceptibility show signs of positive selection, while only 22 associated with disease protection show signs of positive selection. Alleles increasing susceptibility to RA are under selection as well. In contrast, selection in SNPs associated with CD favors protective alleles. These results inform the current understanding of disease etiology, shed light on potential benefits associated with the genetic-basis of disease, and aid in the efforts to identify causal genetic factors underlying complex disease.

Autoimmune disease classification by inverse association with SNP alleles

With multiple genome-wide association studies (GWAS) performed across autoimmune diseases, there is a great opportunity to study the homogeneity of genetic architectures across autoimmune disease. Previous approaches have been limited in the scope of their analysis and have failed to properly incorporate the direction of allele-specific disease associations for SNPs. In this work, we refine the notion of a genetic variation profile for a given disease to capture strength of association with multiple SNPs in an allele-specific fashion. We apply this method to compare genetic variation profiles of six autoimmune diseases: multiple sclerosis (MS), ankylosing spondylitis (AS), autoimmune thyroid disease (ATD), rheumatoid arthritis (RA), Crohn's disease (CD), and type 1 diabetes (T1D), as well as five non-autoimmune diseases. We quantify pair-wise relationships between these diseases and find two broad clusters of autoimmune disease where SNPs that make an individual susceptible to one class of autoimmune disease also protect from diseases in the other autoimmune class. We find that RA and AS form one such class, and MS and ATD another. We identify specific SNPs and genes with opposite risk profiles for these two classes. We furthermore explore individual SNPs that play an important role in defining similarities and differences between disease pairs. We present a novel, systematic, cross-platform approach to identify allele-specific relationships between disease pairs based on genetic variation as well as the individual SNPs which drive the relationships. While recognizing similarities between diseases might lead to identifying novel treatment options, detecting differences between diseases previously thought to be similar may point to key novel disease-specific genes and pathways.