Comparison of clustering methods for investigation of genome-wide methylation array data

DNA Methylation and Prospects for Predicting the Therapeutic Effect of Neoadjuvant Chemotherapy for Triple-Negative and Luminal B Breast Cancer

Despite advances in the diagnosis and treatment of breast cancer (BC), the main cause of deaths is resistance to existing therapies. An approach to improve the effectiveness of therapy in patients with aggressive BC subtypes is neoadjuvant chemotherapy (NACT). Yet, the response to NACT for aggressive subtypes is less than 65% according to large clinical trials. An obvious fact is the lack of biomarkers predicting the therapeutic effect of NACT. In a search for epigenetic markers, we performed genome-wide differential methylation screening by XmaI-RRBS in cohorts of NACT responders and nonresponders, for triple-negative (TN) and luminal B tumors. The predictive potential of the most discriminative loci was further assessed in independent cohorts by methylation-sensitive restriction enzyme quantitative PCR (MSRE-qPCR), a promising method for the implementation of DNA methylation markers in diagnostic laboratories. The selected most informative individual markers were combined into panels demonstrating cvAUC = 0.83 (TMEM132D and MYO15B markers panel) for TN tumors and cvAUC = 0.76 (TTC34, LTBR and CLEC14A) for luminal B tumors. The combination of methylation markers with clinical features that correlate with NACT effect (clinical stage for TN and lymph node status for luminal B tumors) produces better classifiers, with cvAUC = 0.87 for TN tumors and cvAUC = 0.83 for luminal B tumors. Thus, clinical characteristics predictive of NACT response are independently additive to the epigenetic classifier and in combination improve prediction.

Investigating data-driven biological subtypes of psychiatric disorders using specification-curve analysis

BACKGROUND

Cluster analyses have become popular tools for data-driven classification in biological psychiatric research. However, these analyses are known to be sensitive to the chosen methods and/or modelling options, which may hamper generalizability and replicability of findings. To gain more insight into this problem, we used Specification-Curve Analysis (SCA) to investigate the influence of methodological variation on biomarker-based cluster-analysis results.

METHODS

Proteomics data (31 biomarkers) were used from patients (n = 688) and healthy controls (n = 426) in the Netherlands Study of Depression and Anxiety. In SCAs, consistency of results was evaluated across 1200 k-means and hierarchical clustering analyses, each with a unique combination of the clustering algorithm, fit-index, and distance metric. Next, SCAs were run in simulated datasets with varying cluster numbers and noise/outlier levels to evaluate the effect of data properties on SCA outcomes.

RESULTS

The real data SCA showed no robust patterns of biological clustering in either the MDD or a combined MDD/healthy dataset. The simulation results showed that the correct number of clusters could be identified quite consistently across the 1200 model specifications, but that correct cluster identification became harder when the number of clusters and noise levels increased.

CONCLUSION

SCA can provide useful insights into the presence of clusters in biomarker data. However, SCA is likely to show inconsistent results in real-world biomarker datasets that are complex and contain considerable levels of noise. Here, the number and nature of the observed clusters may depend strongly on the chosen model-specification, precluding conclusions about the existence of biological clusters among psychiatric patients.

Effect of prenatal exposure to phthalates on epigenome-wide DNA methylations in cord blood and implications for fetal growth: The Hokkaido Study on Environment and Children's Health

Prenatal exposure to phthalates negatively affects the offspring's health. In particular, epigenetic alterations, such as DNA methylation, may connect phthalate exposure with health outcomes. Here, we evaluated the association of di-2-ethylhexyl phthalate (DEHP) exposure in utero with cord blood epigenome-wide DNA methylation in 203 mother-child pairs enrolled in the Hokkaido Study on Environment and Children's Health, using the Illumina HumanMethylation450 BeadChip. Epigenome-wide association analysis demonstrated the predominant positive associations between the levels of the primary metabolite of DEHP, mono(2-ethylhexyl) phthalate (MEHP), in maternal blood and DNA methylation levels in cord blood. The genes annotated to the CpGs positively associated with MEHP levels were enriched for pathways related to metabolism, the endocrine system, and signal transduction. Among them, methylation levels of CpGs involved in metabolism were inversely associated with the offspring's ponderal index (PI). Further, clustering and mediation analyses suggested that multiple increased methylation changes may jointly mediate the association of DEHP exposure in utero with the offspring's PI at birth. Although further studies are required to assess the impact of these changes, this study suggests that differential DNA methylation may link phthalate exposure in utero to fetal growth and further imply that DNA methylation has predictive value for the offspring's obesity.

iDEP Web Application for RNA-Seq Data Analysis

RNA sequencing (RNA-seq) has become a routine method for transcriptomic profiling. We developed a user-friendly web app called iDEP (integrated differential expression and pathway analysis) to help biologists interpret read counts or other types of expression matrices derived from read mapping. With iDEP, users can easily conduct exploratory data analysis, identify differentially expressed genes, and perform pathway analysis. Due to its intuitive user interface and massive annotation database, iDEP is being widely adopted for interactive analysis of RNA-seq data. Using a public dataset on the effect of heat shock on mouse with and without functional Hsf1, we demonstrate how users can prepare data files and conduct in-depth analysis. We also discuss the importance of critical interpretion of results (avoid p-hacking and rationalizing) and validation of significant pathways by using different methods and independent annotation databases.

Differential methylation of G-protein coupled receptor signaling genes in gastrointestinal neuroendocrine tumors

Neuroendocrine tumors (NETs) of the small intestine undergo large chromosomal and methylation changes. The objective of this study was to identify methylation differences in NETs and consider how the differentially methylated genes may impact patient survival. Genome-wide methylation and chromosomal copy number variation (CNV) of NETs from the small intestine and appendix were measured. Tumors were divided into three molecular subtypes according to CNV results: chromosome 18 loss (18LOH), Multiple CNV, and No CNV. Comparison of 18LOH tumors with MultiCNV and NoCNV tumors identified 901 differentially methylated genes. Genes from the G-protein coupled receptor (GPCR) pathways are statistically overrepresented in the differentially methylated genes. One of the highlighted genes from the GPCR pathway is somatostatin (SST), a clinical target for NETs. Patient survival based on low versus high methylation in all samples identified four significant genes (p < 0.05) OR2S2, SMILR, RNU6-653P, and AC010543.1. Within the 18LOH molecular subtype tumors, survival differences were identified in high versus low methylation of 24 genes. The most significant is TRHR (p < 0.01), a GPCR with multiple FDA-approved drugs. By separating NETs into different molecular subtypes based on chromosomal changes, we find that multiple GPCRs and their ligands appear to be regulated through methylation and correlated with survival. These results suggest opportunities for better treatment strategies for NETs based on molecular features.

Comparative analysis of statistical tools for oil palm phytochemical research

Recent advances in phytochemical analysis have allowed the accumulation of data for crop researchers due to its capacity to footprint and distinguish metabolites that are present within an organisms, tissues or cells. Apart from genotypic traits, slight changes either by biotic or abiotic stimuli will have significant impact on the metabolite abundances and will eventually be observed through physicochemical characteristics. Apposite data mining to interpret the mounds of phytochemical information from such a dynamic system is thus incumbent. In this investigation, several statistical software platforms ranging from exploratory and confirmatory technique of multivariate data analysis from four different statistical tools of COVAIN, SIMCA-P+, MetaboAnalyst and RIKEN Excel Macro were appraised using an oil palm phytochemical data set. As different software tool encompasses its own advantages and limitations, the insights gained from this assessment were documented to enlighten several aspects of functions and suitability for the adaptation of the tools into the oil palm phytochemistry pipeline. This comparative analysis will certainly provide scientists with salient notes on data assessment and data mining that will later allow the depiction of the overall oil palm status in-situ and ex-situ.

Transcriptomic stratification of late-onset Alzheimer's cases reveals novel genetic modifiers of disease pathology

Late-Onset Alzheimer's disease (LOAD) is a common, complex genetic disorder well-known for its heterogeneous pathology. The genetic heterogeneity underlying common, complex diseases poses a major challenge for targeted therapies and the identification of novel disease-associated variants. Case-control approaches are often limited to examining a specific outcome in a group of heterogenous patients with different clinical characteristics. Here, we developed a novel approach to define relevant transcriptomic endophenotypes and stratify decedents based on molecular profiles in three independent human LOAD cohorts. By integrating post-mortem brain gene co-expression data from 2114 human samples with LOAD, we developed a novel quantitative, composite phenotype that can better account for the heterogeneity in genetic architecture underlying the disease. We used iterative weighted gene co-expression network analysis (WGCNA) to reduce data dimensionality and to isolate gene sets that are highly co-expressed within disease subtypes and represent specific molecular pathways. We then performed single variant association testing using whole genome-sequencing data for the novel composite phenotype in order to identify genetic loci that contribute to disease heterogeneity. Distinct LOAD subtypes were identified for all three study cohorts (two in ROSMAP, three in Mayo Clinic, and two in Mount Sinai Brain Bank). Single variant association analysis identified a genome-wide significant variant in TMEM106B (p-value < 5×10-8, rs1990620G) in the ROSMAP cohort that confers protection from the inflammatory LOAD subtype. Taken together, our novel approach can be used to stratify LOAD into distinct molecular subtypes based on affected disease pathways.

Disparity between Inter-Patient Molecular Heterogeneity and Repertoires of Target Drugs Used for Different Types of Cancer in Clinical Oncology

Inter-patient molecular heterogeneity is the major declared driver of an expanding variety of anticancer drugs and personalizing their prescriptions. Here, we compared interpatient molecular heterogeneities of tumors and repertoires of drugs or their molecular targets currently in use in clinical oncology. We estimated molecular heterogeneity using genomic (whole exome sequencing) and transcriptomic (RNA sequencing) data for 4890 tumors taken from The Cancer Genome Atlas database. For thirteen major cancer types, we compared heterogeneities at the levels of mutations and gene expression with the repertoires of targeted therapeutics and their molecular targets accepted by the current guidelines in oncology. Totally, 85 drugs were investigated, collectively covering 82 individual molecular targets. For the first time, we showed that the repertoires of molecular targets of accepted drugs did not correlate with molecular heterogeneities of different cancer types. On the other hand, we found that the clinical recommendations for the available cancer drugs were strongly congruent with the gene expression but not gene mutation patterns. We detected the best match among the drugs usage recommendations and molecular patterns for the kidney, stomach, bladder, ovarian and endometrial cancers. In contrast, brain tumors, prostate and colorectal cancers showed the lowest match. These findings provide a theoretical basis for reconsidering usage of targeted therapeutics and intensifying drug repurposing efforts.

Monozygotic twins and triplets discordant for amyotrophic lateral sclerosis display differential methylation and gene expression

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterised by the loss of upper and lower motor neurons. ALS exhibits high phenotypic variability including age and site of onset, and disease duration. To uncover epigenetic and transcriptomic factors that may modify an ALS phenotype, we used a cohort of Australian monozygotic twins (n = 3 pairs) and triplets (n = 1 set) that are discordant for ALS and represent sporadic ALS and the two most common types of familial ALS, linked to C9orf72 and SOD1. Illumina Infinium HumanMethylation450K BeadChip, EpiTYPER and RNA-Seq analyses in these ALS-discordant twins/triplets and control twins (n = 2 pairs), implicated genes with consistent longitudinal differential DNA methylation and/or gene expression. Two identified genes, RAD9B and C8orf46, showed significant differential methylation in an extended cohort of >1000 ALS cases and controls. Combined longitudinal methylation-transcription analysis within a single twin set implicated CCNF, DPP6, RAMP3, and CCS, which have been previously associated with ALS. Longitudinal transcriptome data showed an 8-fold enrichment of immune function genes and under-representation of transcription and protein modification genes in ALS. Examination of these changes in a large Australian sporadic ALS cohort suggest a broader role in ALS. Furthermore, we observe that increased methylation age is a signature of ALS in older patients.

Genome-wide methylotyping resolves breast cancer epigenetic heterogeneity and suggests novel therapeutic perspectives

Aim: To provide a breast cancer (BC) methylotype classification by genome-wide CpG islands bisulfite DNA sequencing. Materials & methods: XmaI-reduced representation bisulfite sequencing DNA methylation sequencing method was used to profile DNA methylation of 110 BC samples and 6 normal breast samples. Intrinsic DNA methylation BC subtypes were elicited by unsupervised hierarchical cluster analysis, and cluster-specific differentially methylated genes were identified. Results & conclusion: Overall, six distinct BC methylotypes were identified. BC cell lines constitute a separate group extremely highly methylated at the CpG islands. In turn, primary BC samples segregate into two major subtypes, highly and moderately methylated. Highly and moderately methylated superclusters, each incorporate three distinct epigenomic BC clusters with specific features, suggesting novel perspectives for personalized therapy.

Endoflip vs high-definition manometry in the assessment of fecal incontinence: A data-driven unsupervised comparison

BACKGROUND

How much anal sphincter dysfunction contributes to fecal incontinence (FI) is not clear. High-definition anorectal manometry (HDAM) and functional luminal imaging probe (Endoflip) are two new techniques to study anal sphincter function.

AIMS

The goal was to compare the diagnostic utility of HDAM and Endoflip using optimal feature(s) in each modality for FI diagnosis.

METHODS

Blinded classification was carried out on 70 female subjects (32 FI & 38 controls), using 3 prominent machine-learning clustering techniques, with 3 distance metrics. For HDAM, descriptive statistics, shape, and textural features characterizing the spatial relationship of pixels in the HDAM high-pressure zone, and for Endoflip, permutations of pressure and CSA combinations (ie, multiplication, division, or individually) at rest and squeeze were tested.

RESULTS

Intramodality: (a) Endoflip: Best clustering was obtained using the combination of the ratio of CSA over pressure at 40 and 50 mL at rest, which had significantly better specificity (P < 0.001) than using only pressure at 50 mL, no difference in sensitivity (P = 0.68). (b) HDAM: clustering using textural information at rest had significantly higher specificity compared to using only the maximal pressure at rest (P < 0.001). Intermodality: Clustering results using optimal features were not significantly different with respect to sensitivity or specificity (P > 0.05). Optimal Endoflip feature set differed significantly in specificity compared to HDAM maximal pressure at both rest (P < 0.001) and squeeze (P < 0.001).

CONCLUSION

Defective anal closure function is fairly sensitive and highly specific in diagnosing FI. Using optimal feature sets, HDAM and Endoflip perform in a similar fashion in diagnosing FI, but are not complementary.

Creating functional groups of marine fish from categorical traits

Background

Functional groups serve two important functions in ecology: they allow for simplification of ecosystem models and can aid in understanding diversity. Despite their important applications, there has not been a universally accepted method of how to define them. A common approach is to cluster species on a set of traits, validated through visual confirmation of resulting groups based primarily on expert opinion. The goal of this research is to determine a suitable procedure for creating and evaluating functional groups that arise from clustering nominal traits.

Methods

To do so, we produced a species by trait matrix of 22 traits from 116 fish species from Tasman Bay and Golden Bay, New Zealand. Data collected from photographs and published literature were predominantly nominal, and a small number of continuous traits were discretized. Some data were missing, so the benefit of imputing data was assessed using four approaches on data with known missing values. Hierarchical clustering is utilised to search for underlying data structure in the data that may represent functional groups. Within this clustering paradigm there are a number of distance matrices and linkage methods available, several combinations of which we test. The resulting clusters are evaluated using internal metrics developed specifically for nominal clustering. This revealed the choice of number of clusters, distance matrix and linkage method greatly affected the overall within- and between- cluster variability. We visualise the clustering in two dimensions and the stability of clusters is assessed through bootstrapping.

Results

Missing data imputation showed up to 90% accuracy using polytomous imputation, so was used to impute the real missing data. A division of the species information into three functional groups was the most separated, compact and stable result. Increasing the number of clusters increased the inconsistency of group membership, and selection of the appropriate distance matrix and linkage method improved the fit.

Discussion

We show that the commonly used methodologies used for the creation of functional groups are fraught with subjectivity, ultimately causing significant variation in the composition of resulting groups. Depending on the research goal dictates the appropriate strategy for selecting number of groups, distance matrix and clustering algorithm combination.

A systems-level framework for drug discovery identifies Csf1R as an anti-epileptic drug target

The identification of drug targets is highly challenging, particularly for diseases of the brain. To address this problem, we developed and experimentally validated a general computational framework for drug target discovery that combines gene regulatory information with causal reasoning ("Causal Reasoning Analytical Framework for Target discovery"-CRAFT). Using a systems genetics approach and starting from gene expression data from the target tissue, CRAFT provides a predictive framework for identifying cell membrane receptors with a direction-specified influence over disease-related gene expression profiles. As proof of concept, we applied CRAFT to epilepsy and predicted the tyrosine kinase receptor Csf1R as a potential therapeutic target. The predicted effect of Csf1R blockade in attenuating epilepsy seizures was validated in three pre-clinical models of epilepsy. These results highlight CRAFT as a systems-level framework for target discovery and suggest Csf1R blockade as a novel therapeutic strategy in epilepsy. CRAFT is applicable to disease settings other than epilepsy.

An unsupervised machine learning method for discovering patient clusters based on genetic signatures

INTRODUCTION

Many chronic disorders have genomic etiology, disease progression, clinical presentation, and response to treatment that vary on a patient-to-patient basis. Such variability creates a need to identify characteristics within patient populations that have clinically relevant predictive value in order to advance personalized medicine. Unsupervised machine learning methods are suitable to address this type of problem, in which no a priori class label information is available to guide this search. However, it is challenging for existing methods to identify cluster memberships that are not just a result of natural sampling variation. Moreover, most of the current methods require researchers to provide specific input parameters a priori.

METHOD

This work presents an unsupervised machine learning method to cluster patients based on their genomic makeup without providing input parameters a priori. The method implements internal validity metrics to algorithmically identify the number of clusters, as well as statistical analyses to test for the significance of the results. Furthermore, the method takes advantage of the high degree of linkage disequilibrium between single nucleotide polymorphisms. Finally, a gene pathway analysis is performed to identify potential relationships between the clusters in the context of known biological knowledge.

DATASETS AND RESULTS

The method is tested with a cluster validation and a genomic dataset previously used in the literature. Benchmark results indicate that the proposed method provides the greatest performance out of the methods tested. Furthermore, the method is implemented on a sample genome-wide study dataset of 191 multiple sclerosis patients. The results indicate that the method was able to identify genetically distinct patient clusters without the need to select parameters a priori. Additionally, variants identified as significantly different between clusters are shown to be enriched for protein-protein interactions, especially in immune processes and cell adhesion pathways, via Gene Ontology term analysis.

CONCLUSION

Once links are drawn between clusters and clinically relevant outcomes, Immunochip data can be used to classify high-risk and newly diagnosed chronic disease patients into known clusters for predictive value. Further investigation can extend beyond pathway analysis to evaluate these clusters for clinical significance of genetically related characteristics such as age of onset, disease course, heritability, and response to treatment.

Expression profile of desiccation tolerance factors in intertidal seaweed species during the tidal cycle

The transcriptional modulation of desiccation tolerance factors in P. orbicularis explains its successful recuperation after water deficit. Differential responses to air exposure clarify seaweed distribution along intertidal rocky zones. Desiccation-tolerant seaweed species, such as Pyropia orbicularis, can tolerate near 96% water loss during air exposure. To understand the phenotypic plasticity of P. orbicularis to desiccation, several tolerance factors were assessed by RT-qPCR, Western-blot analysis, and enzymatic assays during the natural desiccation-rehydration cycle. Comparative enzymatic analyses were used to evidence differential responses between P. orbicularis and desiccation-sensitive species. The results showed that during desiccation, the relative mRNA levels of genes associated with basal metabolism [trehalose phosphate synthase (tps) and pyruvate dehydrogenase (pdh)] were overexpressed in P. orbicularis. Transcript levels related to antioxidant metabolism [peroxiredoxin (prx); thioredoxin (trx); catalase (cat); lipoxygenase (lox); ferredoxin (fnr); glutathione S-transferase (gst)], cellular detoxification [ABC transporter (abc) and ubiquitin (ubq)], and signal transduction [calmodulin (cam)] increased approximately 15- to 20-fold, with the majority returning to basal levels during the final hours of rehydration. In contrast, actin (act) and transcription factor 1 (tf1) transcripts were down-regulated. ABC transporter protein levels increased in P. orbicularis during desiccation, whereas PRX transcripts decreased. The antioxidant enzymes showed higher specific activity in P. orbicularis under desiccation, and sensitive species exhibited enzymatic inactivation and scarce ABC and PRX protein detection following prolonged desiccation. In conclusion, the reported findings contribute towards understanding the ecological distribution of intertidal seaweeds at the molecular and functional levels.

Single-Cell Transcript Profiles Reveal Multilineage Priming in Early Progenitors Derived from Lgr5(+) Intestinal Stem Cells

Lgr5(+) intestinal stem cells (ISCs) drive epithelial self-renewal, and their immediate progeny-intestinal bipotential progenitors-produce absorptive and secretory lineages via lateral inhibition. To define features of early transit from the ISC compartment, we used a microfluidics approach to measure selected stem- and lineage-specific transcripts in single Lgr5(+) cells. We identified two distinct cell populations, one that expresses known ISC markers and a second, abundant population that simultaneously expresses markers of stem and mature absorptive and secretory cells. Single-molecule mRNA in situ hybridization and immunofluorescence verified expression of lineage-restricted genes in a subset of Lgr5(+) cells in vivo. Transcriptional network analysis revealed that one group of Lgr5(+) cells arises from the other and displays characteristics expected of bipotential progenitors, including activation of Notch ligand and cell-cycle-inhibitor genes. These findings define the earliest steps in ISC differentiation and reveal multilineage gene priming as a fundamental property of the process.

Alcohol and the methylome: design and analysis considerations for research using human samples

BACKGROUND

A growing number of studies in human samples have sought to determine whether chronic alcohol use and alcohol use disorders (AUDs) may be associated with epigenetic factors, such as DNA methylation. We review the extant literature in light of some of the challenges that currently affect the design and interpretation of epigenetic research in human samples.

METHOD

A literature search was used to identify studies that have examined DNA methylation in relation to alcohol use or AUDs in human samples (through July 2013). A total of 22 studies were identified.

RESULTS

Associations with quantitative or diagnostic phenotypes of alcohol use or AUDs have been reported for several genes. However, all studies to date have relied on relatively small samples and cross-sectional study designs. Additionally, attempts to replicate results have been rare. More generally, research progress is hampered by several issues, including limitations of the technologies used to assess DNA methylation, tissue- and cell-specificity of methylation patterns, the difficulties of relating observed methylation differences at a given locus to a functional effect, and limited knowledge about the molecular mechanisms underlying the effects of alcohol on DNA methylation.

CONCLUSIONS

Although we share the optimism that epigenetics may lead to new insights into the etiology and pathophysiology of AUDs, the methodological and scientific challenges associated with conducting methylomic research in human samples need to be carefully considered when designing and evaluating such studies.

Environmental epigenetics: prospects for studying epigenetic mediation of exposure-response relationships

Changes in epigenetic marks such as DNA methylation and histone acetylation are associated with a broad range of disease traits, including cancer, asthma, metabolic disorders, and various reproductive conditions. It seems plausible that changes in epigenetic state may be induced by environmental exposures such as malnutrition, tobacco smoke, air pollutants, metals, organic chemicals, other sources of oxidative stress, and the microbiome, particularly if the exposure occurs during key periods of development. Thus, epigenetic changes could represent an important pathway by which environmental factors influence disease risks, both within individuals and across generations. We discuss some of the challenges in studying epigenetic mediation of pathogenesis and describe some unique opportunities for exploring these phenomena.