A DNA methylation network interaction measure, and detection of network oncomarkers

Parenclitic and Synolytic Networks Revisited

Parenclitic networks provide a powerful and relatively new way to coerce multidimensional data into a graph form, enabling the application of graph theory to evaluate features. Different algorithms have been published for constructing parenclitic networks, leading to the question-which algorithm should be chosen? Initially, it was suggested to calculate the weight of an edge between two nodes of the network as a deviation from a linear regression, calculated for a dependence of one of these features on the other. This method works well, but not when features do not have a linear relationship. To overcome this, it was suggested to calculate edge weights as the distance from the area of most probable values by using a kernel density estimation. In these two approaches only one class (typically controls or healthy population) is used to construct a model. To take account of a second class, we have introduced synolytic networks, using a boundary between two classes on the feature-feature plane to estimate the weight of the edge between these features. Common to all these approaches is that topological indices can be used to evaluate the structure represented by the graphs. To compare these network approaches alongside more traditional machine-learning algorithms, we performed a substantial analysis using both synthetic data with a priori known structure and publicly available datasets used for the benchmarking of ML-algorithms. Such a comparison has shown that the main advantage of parenclitic and synolytic networks is their resistance to over-fitting (occurring when the number of features is greater than the number of subjects) compared to other ML approaches. Secondly, the capability to visualise data in a structured form, even when this structure is not a priori available allows for visual inspection and the application of well-established graph theory to their interpretation/application, eliminating the "black-box" nature of other ML approaches.

MethylSPWNet and MethylCapsNet: Biologically Motivated Organization of DNAm Neural Networks, Inspired by Capsule Networks

DNA methylation (DNAm) alterations have been heavily implicated in carcinogenesis and the pathophysiology of diseases through upstream regulation of gene expression. DNAm deep-learning approaches are able to capture features associated with aging, cell type, and disease progression, but lack incorporation of prior biological knowledge. Here, we present modular, user-friendly deep-learning methodology and software, MethylCapsNet and MethylSPWNet, that group CpGs into biologically relevant capsules-such as gene promoter context, CpG island relationship, or user-defined groupings-and relate them to diagnostic and prognostic outcomes. We demonstrate these models' utility on 3,897 individuals in the classification of central nervous system (CNS) tumors. MethylCapsNet and MethylSPWNet provide an opportunity to increase DNAm deep-learning analyses' interpretability by enabling a flexible organization of DNAm data into biologically relevant capsules.

Clinical epigenetics settings for cancer and cardiovascular diseases: real-life applications of network medicine at the bedside

Despite impressive efforts invested in epigenetic research in the last 50 years, clinical applications are still lacking. Only a few university hospital centers currently use epigenetic biomarkers at the bedside. Moreover, the overall concept of precision medicine is not widely recognized in routine medical practice and the reductionist approach remains predominant in treating patients affected by major diseases such as cancer and cardiovascular diseases. By its' very nature, epigenetics is integrative of genetic networks. The study of epigenetic biomarkers has led to the identification of numerous drugs with an increasingly significant role in clinical therapy especially of cancer patients. Here, we provide an overview of clinical epigenetics within the context of network analysis. We illustrate achievements to date and discuss how we can move from traditional medicine into the era of network medicine (NM), where pathway-informed molecular diagnostics will allow treatment selection following the paradigm of precision medicine.

The Human Body as a Super Network: Digital Methods to Analyze the Propagation of Aging

Biological aging is a complex process involving multiple biological processes. These can be understood theoretically though considering them as individual networks-e.g., epigenetic networks, cell-cell networks (such as astroglial networks), and population genetics. Mathematical modeling allows the combination of such networks so that they may be studied in unison, to better understand how the so-called "seven pillars of aging" combine and to generate hypothesis for treating aging as a condition at relatively early biological ages. In this review, we consider how recent progression in mathematical modeling can be utilized to investigate aging, particularly in, but not exclusive to, the context of degenerative neuronal disease. We also consider how the latest techniques for generating biomarker models for disease prediction, such as longitudinal analysis and parenclitic analysis can be applied to as both biomarker platforms for aging, as well as to better understand the inescapable condition. This review is written by a highly diverse and multi-disciplinary team of scientists from across the globe and calls for greater collaboration between diverse fields of research.

Data Integration in Poplar: 'Omics Layers and Integration Strategies

Populus trichocarpa is an important biofuel feedstock that has been the target of extensive research and is emerging as a model organism for plants, especially woody perennials. This research has generated several large ‘omics datasets. However, only few studies in Populus have attempted to integrate various data types. This review will summarize various ‘omics data layers, focusing on their application in Populus species. Subsequently, network and signal processing techniques for the integration and analysis of these data types will be discussed, with particular reference to examples in Populus.

Multi-network approach to identify differentially methylated gene communities in cancer

BACKGROUND AND OBJECTIVE

High-throughput Next Generation Sequencing tools have generated immense quantity of genome-wide methylation and expression profiling data, resulting in an unprecedented opportunity to unravel the epigenetic regulatory mechanisms underlying cancer. Identifying differentially methylated regions within gene networks is an important step towards revealing the cancer epigenome blueprint. Approaches that integrate gene methylation and expression profiles assume their negative correlation and build a single scaffold network to cluster. However, the exact regulatory mechanism between gene expression and methylation is not precisely deciphered.

METHODS

A consensus-based clustering framework, namely, Differentially Methylated Gene Communities based on Multi-network (DMGC-M) is proposed, that takes multiple gene networks as input and builds a community structure out of evidences from all network types.

RESULTS

Experiments on six cancer datasets from The Cancer Genome Atlas (TCGA) reveal that multi-network approaches produce more discriminative gene communities than integrated approaches.

CONCLUSION

The proposed method will be useful to a number of researchers who seek to identify epigenetic dysregulations in pathways or molecular networks. The findings can also advance recent research efforts in Molecular Pathologic Epidemiology.

Network assessment of demethylation treatment in melanoma: Differential transcriptome-methylome and antigen profile signatures

Background

In melanoma, like in other cancers, both genetic alterations and epigenetic underlie the metastatic process. These effects are usually measured by changes in both methylome and transcriptome profiles, whose cross-correlation remains uncertain. We aimed to assess at systems scale the significance of epigenetic treatment in melanoma cells with different metastatic potential.

Methods and findings

Treatment by DAC demethylation with 5-Aza-2’-deoxycytidine of two melanoma cell lines endowed with different metastatic potential, SKMEL-2 and HS294T, was performed and high-throughput coupled RNA-Seq and RRBS-Seq experiments delivered differential profiles (DiP) of both transcriptomes and methylomes. Methylation levels measured at both TSS and gene body were studied to inspect correlated patterns with wide-spectrum transcript abundance levels quantified in both protein coding and non-coding RNA (ncRNA) regions. The DiP were then mapped onto standard bio-annotation sources (pathways, biological processes) and network configurations were obtained. The prioritized associations for target identification purposes were expected to elucidate the reprogramming dynamics induced by the epigenetic therapy. The interactomic connectivity maps of each cell line were formed to support the analysis of epigenetically re-activated genes. i.e. those supposedly silenced by melanoma. In particular, modular protein interaction networks (PIN) were used, evidencing a limited number of shared annotations, with an example being MAPK13 (cascade of cellular responses evoked by extracellular stimuli). This gene is also a target associated to the PANDAR ncRNA, therapeutically relevant because of its aberrant expression observed in various cancers. Overall, the non-metastatic SKMEL-2 map reveals post-treatment re-activation of a richer pathway landscape, involving cadherins and integrins as signatures of cell adhesion and proliferation. Relatively more lncRNAs were also annotated, indicating more complex regulation patterns in view of target identification. Finally, the antigen maps matched to DiP display other differential signatures with respect to the metastatic potential of the cell lines. In particular, as demethylated melanomas show connected targets that grow with the increased metastatic potential, also the potential target actionability seems to depend to some degree on the metastatic state. However, caution is required when assessing the direct influence of re-activated genes over the identified targets. In light of the stronger treatment effects observed in non-metastatic conditions, some limitations likely refer to in silico data integration tools and resources available for the analysis of tumor antigens.

Conclusion

Demethylation treatment strongly affects early melanoma progression by re-activating many genes. This evidence suggests that the efficacy of this type of therapeutic intervention is potentially high at the pre-metastatic stages. The biomarkers that can be assessed through antigens seem informative depending on the metastatic conditions, and networks help to elucidate the assessment of possible targets actionability.

Identification of prognostic signature in cancer based on DNA methylation interaction network

BACKGROUND

The identification of prognostic biomarkers for cancer patients is essential for cancer research. These days, DNA methylation has been proved to be associated with cancer prognosis. However, there are few methods which identify the prognostic markers based on DNA methylation data systematically, especially considering the interaction among DNA methylation sites.

METHODS

In this paper, we first evaluated the stabilities of microRNA, mRNA, and DNA methylation data in prognosis of cancer. After that, a rank-based method was applied to construct a DNA methylation interaction network. In this network, nodes with the largest degrees (10% of all the nodes) were selected as hubs. Cox regression was applied to select the hubs as prognostic signature. In this prognostic signature, DNA methylation levels of each DNA methylation site are correlated with the outcomes of cancer patients. After obtaining these prognostic genes, we performed the survival analysis in the training group and the test group to verify the reliability of these genes.

RESULTS

We applied our method in three cancers (ovarian cancer, breast cancer and Glioblastoma Multiforme). In all the three cancers, there are more common ones of prognostic genes selected from different samples in DNA methylation data, compared with gene expression data and miRNA expression data, which indicates the DNA methylation data may be more stable in cancer prognosis. Power-law distribution fitting suggests that the DNA methylation interaction networks are scale-free. And the hubs selected from the three networks are all enriched by cancer related pathways. The gene signatures were obtained for the three cancers respectively, and survival analysis shows they can distinguish the outcomes of tumor patients in both the training data sets and test data sets, which outperformed the control signatures.

CONCLUSIONS

A computational method was proposed to construct DNA methylation interaction network and this network could be used to select prognostic signatures in cancer.

Parenclitic Network Analysis of Methylation Data for Cancer Identification

We make use of ideas from the theory of complex networks to implement a machine learning classification of human DNA methylation data, that carry signatures of cancer development. The data were obtained from patients with various kinds of cancers and represented as parenclictic networks, wherein nodes correspond to genes, and edges are weighted according to pairwise variation from control group subjects. We demonstrate that for the 10 types of cancer under study, it is possible to obtain a high performance of binary classification between cancer-positive and negative samples based on network measures. Remarkably, an accuracy as high as 93−99% is achieved with only 12 network topology indices, in a dramatic reduction of complexity from the original 15295 gene methylation levels. Moreover, it was found that the parenclictic networks are scale-free in cancer-negative subjects, and deviate from the power-law node degree distribution in cancer. The node centrality ranking and arising modular structure could provide insights into the systems biology of cancer.

Intra-Gene DNA Methylation Variability Is a Clinically Independent Prognostic Marker in Women's Cancers

We introduce a novel per-gene measure of intra-gene DNA methylation variability (IGV) based on the Illumina Infinium HumanMethylation450 platform, which is prognostic independently of well-known predictors of clinical outcome. Using IGV, we derive a robust gene-panel prognostic signature for ovarian cancer (OC, n = 221), which validates in two independent data sets from Mayo Clinic (n = 198) and TCGA (n = 358), with significance of p = 0.004 in both sets. The OC prognostic signature gene-panel is comprised of four gene groups, which represent distinct biological processes. We show the IGV measurements of these gene groups are most likely a reflection of a mixture of intra-tumour heterogeneity and transcription factor (TF) binding/activity. IGV can be used to predict clinical outcome in patients individually, providing a surrogate read-out of hard-to-measure disease processes.

Differential analysis of biological networks

Background

In cancer research, the comparison of gene expression or DNA methylation networks inferred from healthy controls and patients can lead to the discovery of biological pathways associated to the disease. As a cancer progresses, its signalling and control networks are subject to some degree of localised re-wiring. Being able to detect disrupted interaction patterns induced by the presence or progression of the disease can lead to the discovery of novel molecular diagnostic and prognostic signatures. Currently there is a lack of scalable statistical procedures for two-network comparisons aimed at detecting localised topological differences.

Results

We propose the dGHD algorithm, a methodology for detecting differential interaction patterns in two-network comparisons. The algorithm relies on a statistic, the Generalised Hamming Distance (GHD), for assessing the degree of topological difference between networks and evaluating its statistical significance. dGHD builds on a non-parametric permutation testing framework but achieves computationally efficiency through an asymptotic normal approximation.

Conclusions

We show that the GHD is able to detect more subtle topological differences compared to a standard Hamming distance between networks. This results in the dGHD algorithm achieving high performance in simulation studies as measured by sensitivity and specificity. An application to the problem of detecting differential DNA co-methylation subnetworks associated to ovarian cancer demonstrates the potential benefits of the proposed methodology for discovering network-derived biomarkers associated with a trait of interest.

Crosstalk analysis of pathways in breast cancer using a network model based on overlapping differentially expressed genes

Multiple signal transduction pathways can affect each other considerably through crosstalk. However, the presence and extent of this phenomenon have not been rigorously studied. The aim of the present study was to identify strong and normal interactions between pathways in breast cancer and determine the main pathway. Five sets of breast cancer data were downloaded from the high-throughput Gene Expression Omnibus (GEO) and analyzed to identify differentially expressed (DE) genes using the Rank Product (RankProd) method. A list of pathways with differential expression was obtained by gene set enrichment analysis (GSEA) of the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. The DE genes that overlapped between pathways were identified and a crosstalk network diagram based on the overlap of DE genes was constructed. A total of 1,464 DE genes and 26 pathways were identified. In addition, the number of DE genes that overlapped between specific pathways were determined, and the greatest degree of overlap was between the extracellular matrix (ECM)-receptor interaction and Focal adhesion pathways, which had 22 overlapping DE genes. Weighted pathway analysis of the crosstalk between pathways identified that Pathways in cancer was the main pathway in breast cancer.