Assessing the impact of HIV treatment interruptions using stochastic cellular Automata.

Chronic HIV infection causes a progressive decrease in the ability to maintain homeostasis resulting, after some time, in eventual break down of immune functions. Recent clinical research has shed light on a significant contribution of the lymphatic tissues, where HIV causes accumulation of collagen, (fibrosis). Specifically, where tissue is populated by certain types of functional stromal cells designated Fibroblastic Reticular Cells (FRCs), these have been found to play a crucial role in balancing out apoptosis and regeneration of naïve T-cells through 2-way cellular signaling. Tissue fibrosis not only impedes this signaling, effectively reducing T-cell levels through increased apoptosis of cells of both T- and FRC type but has been found to be irreversible by current HIV standard treatment (cART). While the therapy aims to block the viral lifecycle, cART-associated increase of T-cell levels in blood appears to conceal existing FRC impairment through fibrosis. This hidden impairment can lead to adverse consequences if treatment is interrupted, e.g. due to poor adherence (missing doses) or through periods recovering from drug toxicities. Formal clinical studies on treatment interruption have indicated possible adverse effects, but quantification of those effects in relation to interruption protocol and patient predisposition remains unclear. Accordingly, the impact of treatment interruption on lymphatic tissue structure and T-cell levels is explored here by means of computer simulation. A novel Stochastic Cellular Automata model is proposed, which utilizes all sources of clinical detail available to us (though sparse in part) for model parametrization. Sources are explicitly referenced and conflicting evidence from previous studies explored. The main focus is on (i) spatial aspects of collagen build up, together with (ii) collagen increase after repeated treatment interruptions to explore the dynamics of HIV-induced fibrosis and T-cell loss.

Colon cancer associated genes exhibit signatures of positive selection at functionally significant positions.

BACKGROUND: Cancer, much like most human disease, is routinely studied by utilizing model organisms. Of these model organisms, mice are often dominant. However, our assumptions of functional equivalence fail to consider the opportunity for divergence conferred by ~180 Million Years (MY) of independent evolution between these species. For a given set of human disease related genes, it is therefore important to determine if functional equivalency has been retained between species. In this study we test the hypothesis that cancer associated genes have different patterns of substitution akin to adaptive evolution in different mammal lineages. RESULTS: Our analysis of the current literature and colon cancer databases identified 22 genes exhibiting colon cancer associated germline mutations. We identified orthologs for these 22 genes across a set of high coverage (>6X) vertebrate genomes. Analysis of these orthologous datasets revealed significant levels of positive selection. Evidence of lineage-specific positive selection was identified in 14 genes in both ancestral and extant lineages. Lineage-specific positive selection was detected in the ancestral Euarchontoglires and Hominidae lineages for STK11, in the ancestral primate lineage for CDH1, in the ancestral Murinae lineage for both SDHC and MSH6 genes and the ancestral Muridae lineage for TSC1. CONCLUSION: Identifying positive selection in the Primate, Hominidae, Muridae and Murinae lineages suggests an ancestral functional shift in these genes between the rodent and primate lineages. Analyses such as this, combining evolutionary theory and predictions - along with medically relevant data, can thus provide us with important clues for modeling human diseases.

Comparison of evolutionary algorithms in gene regulatory network model inference.

BACKGROUND: The evolution of high throughput technologies that measure gene expression levels has created a data base for inferring GRNs (a process also known as reverse engineering of GRNs). However, the nature of these data has made this process very difficult. At the moment, several methods of discovering qualitative causal relationships between genes with high accuracy from microarray data exist, but large scale quantitative analysis on real biological datasets cannot be performed, to date, as existing approaches are not suitable for real microarray data which are noisy and insufficient. RESULTS: This paper performs an analysis of several existing evolutionary algorithms for quantitative gene regulatory network modelling. The aim is to present the techniques used and offer a comprehensive comparison of approaches, under a common framework. Algorithms are applied to both synthetic and real gene expression data from DNA microarrays, and ability to reproduce biological behaviour, scalability and robustness to noise are assessed and compared. CONCLUSIONS: Presented is a comparison framework for assessment of evolutionary algorithms, used to infer gene regulatory networks. Promising methods are identified and a platform for development of appropriate model formalisms is established.

Cell type-dependent, infection-induced, aberrant DNA methylation in gastric cancer.

Epigenetic changes correspond to heritable modifications of the chromatin structure, which do not involve any alteration of the DNA sequence but nonetheless affect gene expression. These mechanisms play an important role in cell differentiation, but aberrant occurrences are also associated with a number of diseases, including cancer and neural development disorders. In particular, aberrant DNA methylation induced by H. Pylori has been found to be a significant risk factor in gastric cancer. To investigate the sensitivity of different genes and cell types to this infection, a computational model of methylation in gastric crypts is developed. In this article, we review existing results from physical experiments and outline their limitations, before presenting the computational model and investigating the influence of its parameters.

Quantitative multi-agent models for simulating protein release from PLGA bioerodible nano- and microspheres.

Using poly(lactide-co-glycolide) (PLGA) particles for drug encapsulation and delivery has recently gained considerable popularity for a number of reasons. An advantage in one sense, but a drawback of PLGA use in another, is that drug delivery systems made of this material can provide a wide range of dissolution profiles, due to their internal structure and properties related to particles' manufacture. The advantages of enriching particulate drug design experimentation with computer models, are evident with simulations used to predict and optimize design, as well as indicate choice of best manufacturing parameters. In the present work, we seek to understand the phenomena observed for PLGA micro- and nanospheres, through Cellular Automata (CA) agent-based Monte Carlo (MC) models. Systems are studied both over large temporal scales (capturing slow erosion of PLGA) and for various spatial configurations (capturing initial as well as dynamic morphology). The major strength of this multi-agent approach is to observe dissolution directly, by monitoring the emergent behaviour: the dissolution profile manifested, as a sphere erodes. Different problematic aspects of the modelling process are discussed in details in this paper. The models were tested on experimental data from literature, demonstrating very good performance. Quantitative discussion is provided throughout the text in order to make a demonstration of the use in practice of the proposed model.

EpiGeNet: A Graph Database of Interdependencies Between Genetic and Epigenetic Events in Colorectal Cancer.

The development of colorectal cancer (CRC)-the third most common cancer type-has been associated with deregulations of cellular mechanisms stimulated by both genetic and epigenetic events. StatEpigen is a manually curated and annotated database, containing information on interdependencies between genetic and epigenetic signals, and specialized currently for CRC research. Although StatEpigen provides a well-developed graphical user interface for information retrieval, advanced queries involving associations between multiple concepts can benefit from more detailed graph representation of the integrated data. This can be achieved by using a graph database (NoSQL) approach. Data were extracted from StatEpigen and imported to our newly developed EpiGeNet, a graph database for storage and querying of conditional relationships between molecular (genetic and epigenetic) events observed at different stages of colorectal oncogenesis. We illustrate the enhanced capability of EpiGeNet for exploration of different queries related to colorectal tumor progression; specifically, we demonstrate the query process for (i) stage-specific molecular events, (ii) most frequently observed genetic and epigenetic interdependencies in colon adenoma, and (iii) paths connecting key genes reported in CRC and associated events. The EpiGeNet framework offers improved capability for management and visualization of data on molecular events specific to CRC initiation and progression.

Computational Modeling and Analysis of Microarray Data: New Horizons.

High-throughput microarray technologies have long been a source of data for a wide range of biomedical investigations. Over the decades, variants have been developed and sophistication of measurements has improved, with generated data providing both valuable insight and considerable analytical challenge. The cost-effectiveness of microarrays, as well as their fundamental applicability, made them a first choice for much early genomic research and efforts to improve accessibility, quality and interpretation have continued unabated. In recent years, however, the emergence of new generations of sequencing methods and, importantly, reduction of costs, has seen a preferred shift in much genomic research to the use of sequence data, both less 'noisy' and, arguably, with species information more directly targeted and easily interpreted. Nevertheless, new microarray data are still being generated and, together with their considerable legacy, can offer a complementary perspective on biological systems and disease pathogenesis. The challenge now is to exploit novel methods for enhancing and combining these data with those generated by alternative high-throughput techniques, such as sequencing, to provide added value. Augmentation and integration of microarray data and the new horizons this opens up, provide the theme for the papers in this Special Issue.

Theoretical cross-comparative analysis on dynamics of small intestine and colon crypts during cancer initiation.

Epigenetics is emerging as a fundamentally important area of biological and medical research that has implications for our understanding of human diseases including cancer, autoimmune and neuropsychiatric disorders. In the context of recent efforts on personalised medicine, a novel research direction is concerned with identification of intra-individual epigenetic variation linked to disease predisposition and development, i.e. epigenome-wide association studies. A computational model has been developed to describe the dynamics and structure of human intestinal crypts and to perform a comparative analysis on aberrant DNA methylation level induced in these during cancer initiation. The crypt framework, AgentCrypt, is an agent-based model of crypt dynamics, which handles intra- and inter-dependencies. In addition, the AgentCrypt model is used to investigate the effect of a set of potential inhibitors with respect to methylation modification in intestinal tissue during initiation of disease. Methylation level decrease over a relatively short period of 90 days is marked for the colon compared to the small intestine, although similar alterations are induced in both tissues. In addition, inhibitor effect is notable for abnormal crypt groups, with largest average methylation differences observed ≈0.75% lower in the colon and ≈0.79% lower in the small intestine with inhibitor present.

Comparative Correlation Structure of Colon Cancer Locus Specific Methylation: Characterisation of Patient Profiles and Potential Markers across 3 Array-Based Datasets.

Abnormal DNA-methylation is well known to play an important role in cancer onset and development, and colon cancer is no exception to this rule. Recent years have seen the increased use of large-scale technologies, (such as methylation microarray assays or specific sequencing of methylated DNA), to determine whole genome profiles of CpG island methylation in tissue samples. Comprehensive study of methylation array data from transcriptome high-throughput platforms permits determination of gene methylation markers, important for cancer profiling. Here, three large-scale methylation datasets for colon cancer have been compared to determine locus-specific methylation agreement. These data are from the GEO database, where colon cancer and apparently healthy adjacent tissues are represented by sample sizes 125 and 29 respectively in the first dataset, 24 of each in the second and 118 of each in the third. Several data analysis techniques have been employed, including Clustering, Discriminant Principal Component Analysis, Discriminant Analysis and ROC curves, in order (i) to obtain a better insight on the locus-specific concomitant methylation structures for these diverse data and (ii) to determine a robust potential marker set for indicative screening, drawn from all data taken together. The extent of the agreement between the analysed datasets is reported. Further, potential screening methylation markers, for which methylation profiles are consistent across tissue samples and several datasets, are highlighted and discussed.

Data Integration for Microarrays: Enhanced Inference for Gene Regulatory Networks.

Microarray technologies have been the basis of numerous important findings regarding gene expression in the few last decades. Studies have generated large amounts of data describing various processes, which, due to the existence of public databases, are widely available for further analysis. Given their lower cost and higher maturity compared to newer sequencing technologies, these data continue to be produced, even though data quality has been the subject of some debate. However, given the large volume of data generated, integration can help overcome some issues related, e.g., to noise or reduced time resolution, while providing additional insight on features not directly addressed by sequencing methods. Here, we present an integration test case based on public Drosophila melanogaster datasets (gene expression, binding site affinities, known interactions). Using an evolutionary computation framework, we show how integration can enhance the ability to recover transcriptional gene regulatory networks from these data, as well as indicating which data types are more important for quantitative and qualitative network inference. Our results show a clear improvement in performance when multiple datasets are integrated, indicating that microarray data will remain a valuable and viable resource for some time to come.

Integrating Colon Cancer Microarray Data: Associating Locus-Specific Methylation Groups to Gene Expression-Based Classifications.

Recently, considerable attention has been paid to gene expression-based classifications of colorectal cancers (CRC) and their association with patient prognosis. In addition to changes in gene expression, abnormal DNA-methylation is known to play an important role in cancer onset and development, and colon cancer is no exception to this rule. Large-scale technologies, such as methylation microarray assays and specific sequencing of methylated DNA, have been used to determine whole genome profiles of CpG island methylation in tissue samples. In this article, publicly available microarray-based gene expression and methylation data sets are used to characterize expression subtypes with respect to locus-specific methylation. A major objective was to determine whether integration of these data types improves previously characterized subtypes, or provides evidence for additional subtypes. We used unsupervised clustering techniques to determine methylation-based subgroups, which are subsequently annotated with three published expression-based classifications, comprising from three to six subtypes. Our results showed that, while methylation profiles provide a further basis for segregation of certain (Inflammatory and Goblet-like) finer-grained expression-based subtypes, they also suggest that other finer-grained subtypes are not distinctive and can be considered as a single subtype.

Diversity emergence and dynamics during primary immune response: a shape space, physical space model.

A computational model of the dynamics of diversity among T-cell receptors and MHC: peptide complex molecules is presented. We propose a method by which individual immune systems may evolve effcient or ineffcient states as a result of T-cell receptor crossreactivity as well as genetic variation among pathogens. By combining shape space and physical space models, valuable insight is obtained into how immune system-wide state is, in large part, determined by localised space dynamics. In the model, system-wide state also informs local dynamics, especially in the lymphatic system during primary immune response. The process by which similar initial infection conditions across individuals may result in highly variable end states (a phenomenon observed in the clinical context) is modelled. Our results show that activity alone is not a good indicator of infection suppression or removal. In this work, we postulate that successful viral clearance is characterised by broad T-cell receptor activation (in shape space), and results in low average concentration levels of activated cytotoxic lymphocyte cells.

A computational model for genetic and epigenetic signals in colon cancer.

Cancer, a class of diseases, characterized by abnormal cell growth, has one of the highest overall death rates world-wide. Its development has been linked to aberrant genetic and epigenetic events, affecting the regulation of key genes that control cellular mechanisms. However, a major issue in cancer research is the lack of precise information on tumour pathways; therefore, the delineation of these and of the processes underlying disease proliferation is an important area of investigation. A computational approach to modelling malignant system events can help to improve understanding likely "triggers", i.e. initiating abnormal micro-molecular signals that occur during cancer development. Here, we introduce a network-based model for genetic and epigenetic events observed at different stages of colon cancer, with a focus on the gene relationships and tumour pathways. Additionally, we describe a case study on tumour progression recorded for two gene networks on colon cancer, carcinoma in situ. Our results to date showed that tumour progression rate is higher for a small, closely-associated network of genes than for a larger, less-connected set; thus, disease development depends on assessment of network properties. The current work aims to provide improved insight on the way in which aberrant modifications characterize cancer initiation and progression. The framework dynamics are described in terms of interdependencies between three main layers: genetic and epigenetic events, gene relationships and cancer stage levels.

Integrating heterogeneous gene expression data for gene regulatory network modelling.

Gene regulatory networks (GRNs) are complex biological systems that have a large impact on protein levels, so that discovering network interactions is a major objective of systems biology. Quantitative GRN models have been inferred, to date, from time series measurements of gene expression, but at small scale, and with limited application to real data. Time series experiments are typically short (number of time points of the order of ten), whereas regulatory networks can be very large (containing hundreds of genes). This creates an under-determination problem, which negatively influences the results of any inferential algorithm. Presented here is an integrative approach to model inference, which has not been previously discussed to the authors' knowledge. Multiple heterogeneous expression time series are used to infer the same model, and results are shown to be more robust to noise and parameter perturbation. Additionally, a wavelet analysis shows that these models display limited noise over-fitting within the individual datasets.

RNA-Seq vs dual- and single-channel microarray data: sensitivity analysis for differential expression and clustering.

With the fast development of high-throughput sequencing technologies, a new generation of genome-wide gene expression measurements is under way. This is based on mRNA sequencing (RNA-seq), which complements the already mature technology of microarrays, and is expected to overcome some of the latter's disadvantages. These RNA-seq data pose new challenges, however, as strengths and weaknesses have yet to be fully identified. Ideally, Next (or Second) Generation Sequencing measures can be integrated for more comprehensive gene expression investigation to facilitate analysis of whole regulatory networks. At present, however, the nature of these data is not very well understood. In this paper we study three alternative gene expression time series datasets for the Drosophila melanogaster embryo development, in order to compare three measurement techniques: RNA-seq, single-channel and dual-channel microarrays. The aim is to study the state of the art for the three technologies, with a view of assessing overlapping features, data compatibility and integration potential, in the context of time series measurements. This involves using established tools for each of the three different technologies, and technical and biological replicates (for RNA-seq and microarrays, respectively), due to the limited availability of biological RNA-seq replicates for time series data. The approach consists of a sensitivity analysis for differential expression and clustering. In general, the RNA-seq dataset displayed highest sensitivity to differential expression. The single-channel data performed similarly for the differentially expressed genes common to gene sets considered. Cluster analysis was used to identify different features of the gene space for the three datasets, with higher similarities found for the RNA-seq and single-channel microarray dataset.

The landscape for epigenetic/epigenomic biomedical resources.

Recent advances in molecular biology and computational power have seen the biomedical sector enter a new era, with corresponding development of Bioinformatics as a major discipline. Generation of enormous amounts of data has driven the need for more advanced storage solutions and shared access through a range of public repositories. The number of such biomedical resources is increasing constantly and mining these large and diverse data sets continues to present real challenges. This paper attempts a general overview of currently available resources, together with remarks on their data mining and analysis capabilities. Of interest here is the recent shift in focus from genetic to epigenetic/epigenomic research and the emergence and extension of resource provision to support this both at local and global scale. Biomedical text and numerical data mining are both considered, the first dealing with automated methods for analyzing research content and information extraction, and the second (broadly) with pattern recognition and prediction. Any summary and selection of resources is inherently limited, given the spectrum available, but the aim is to provide a guideline for the assessment and comparison of currently available provision, particularly as this relates to epigenetics/epigenomics.

Cross-platform microarray data normalisation for regulatory network inference.

BACKGROUND: Inferring Gene Regulatory Networks (GRNs) from time course microarray data suffers from the dimensionality problem created by the short length of available time series compared to the large number of genes in the network. To overcome this, data integration from diverse sources is mandatory. Microarray data from different sources and platforms are publicly available, but integration is not straightforward, due to platform and experimental differences. METHODS: We analyse here different normalisation approaches for microarray data integration, in the context of reverse engineering of GRN quantitative models. We introduce two preprocessing approaches based on existing normalisation techniques and provide a comprehensive comparison of normalised datasets. CONCLUSIONS: Results identify a method based on a combination of Loess normalisation and iterative K-means as best for time series normalisation for this problem.

Model refinement through high-performance computing: an agent-based HIV example.

BACKGROUND: Recent advances in Immunology highlighted the importance of local properties on the overall progression of HIV infection. In particular, the gastrointestinal tract is seen as a key area during early infection, and the massive cell depletion associated with it may influence subsequent disease progression. This motivated the development of a large-scale agent-based model. RESULTS: Lymph nodes are explicitly implemented, and considerations on parallel computing permit large simulations and the inclusion of local features. The results obtained show that GI tract inclusion in the model leads to an accelerated disease progression, during both the early stages and the long-term evolution, compared to a theoretical, uniform model. CONCLUSIONS: These results confirm the potential of treatment policies currently under investigation, which focus on this region. They also highlight the potential of this modelling framework, incorporating both agent-based and network-based components, in the context of complex systems where scaling-up alone does not result in models providing additional insights.

Computational micromodel for epigenetic mechanisms.

Characterization of the epigenetic profile of humans since the initial breakthrough on the human genome project has strongly established the key role of histone modifications and DNA methylation. These dynamic elements interact to determine the normal level of expression or methylation status of the constituent genes in the genome. Recently, considerable evidence has been put forward to demonstrate that environmental stress implicitly alters epigenetic patterns causing imbalance that can lead to cancer initiation. This chain of consequences has motivated attempts to computationally model the influence of histone modification and DNA methylation in gene expression and investigate their intrinsic interdependency. In this paper, we explore the relation between DNA methylation and transcription and characterize in detail the histone modifications for specific DNA methylation levels using a stochastic approach.