Learning subgroup-specific regulatory interactions and regulator independence with PARADIGM

mosGraphGen: a novel tool to generate multi-omics signaling graphs to facilitate integrative and interpretable graph AI model development

Motivation

Multi-omics data, i.e. genomics, epigenomics, transcriptomics, proteomics, characterize cellular complex signaling systems from multi-level and multi-view and provide a holistic view of complex cellular signaling pathways. However, it remains challenging to integrate and interpret multi-omics data for mining critical biomarkers. Graph AI models have been widely used to analyze graph-structure datasets, and are ideal for integrative multi-omics data analysis because they can naturally integrate and represent multi-omics data as a biologically meaningful multi-level signaling graph and interpret multi-omics data via graph node and edge ranking analysis. Nevertheless, it is nontrivial for graph-AI model developers to pre-analyze multi-omics data and convert the data into biologically meaningful graphs, which can be directly fed into graph-AI models.

Results

To resolve this challenge, we developed mosGraphGen (multi-omics signaling graph generator), generating Multi-omics Signaling graphs (mos-graph) of individual samples by mapping multi-omics data onto a biologically meaningful multi-level background signaling network with data normalization by aggregating measurements and aligning to the reference genome. With mosGraphGen, AI model developers can directly apply and evaluate their models using these mos-graphs. In the results, mosGraphGen was used and illustrated using two widely used multi-omics datasets of The Cancer Genome Atlas (TCGA) and Alzheimer's disease (AD) samples.

Availability and implementation

The code of mosGraphGen is open-source and publicly available via GitHub: https://github.com/FuhaiLiAiLab/mosGraphGen.

mosGraphGen: a novel tool to generate multi-omics signaling graphs to facilitate integrative and interpretable graph AI model development

Multi-omics data, i.e., genomics, epigenomics, transcriptomics, proteomics, characterize cellular complex signaling systems from multi-level and multi-view and provide a holistic view of complex cellular signaling pathways. However, it remains challenging to integrate and interpret multi-omics data for mining key disease targets and signaling pathways. Graph AI models have been widely used to analyze graph-structure datasets, and are ideal for integrative multi-omics data analysis because they can naturally integrate and represent multi-omics data as a biologically meaningful multi-level signaling graph and interpret multi-omics data via graph node and edge ranking analysis. However, it is non-trivial for graph-AI model developers to pre-analyze multi-omics data and convert the data into biologically meaningful graphs, which can be directly fed into graph-AI models. To resolve this challenge, we developed mosGraphGen (multi-omics signaling graph generator), generating Multi-omics Signaling graphs (mos-graph) of individual samples by mapping multi-omics data onto a biologically meaningful multi-level background signaling network with data normalization by aggregating measurements and aligning to the reference genome. With mosGraphGen, AI model developers can directly apply and evaluate their models using these mos-graphs. In the results, mosGraphGen was used and illustrated using two widely used multi-omics datasets of TCGA and Alzheimer's disease (AD) samples. The code of mosGraphGen is open-source and publicly available via GitHub: https://github.com/FuhaiLiAiLab/mosGraphGen.

MPAC: a computational framework for inferring cancer pathway activities from multi-omic data

Fully capturing cellular state requires examining genomic, epigenomic, transcriptomic, proteomic, and other assays for a biological sample and comprehensive computational modeling to reason with the complex and sometimes conflicting measurements. Modeling these so-called multi-omic data is especially beneficial in disease analysis, where observations across omic data types may reveal unexpected patient groupings and inform clinical outcomes and treatments. We present Multi-omic Pathway Analysis of Cancer (MPAC), a computational framework that interprets multi-omic data through prior knowledge from biological pathways. MPAC uses network relationships encoded in pathways using a factor graph to infer consensus activity levels for proteins and associated pathway entities from multi-omic data, runs permutation testing to eliminate spurious activity predictions, and groups biological samples by pathway activities to prioritize proteins with potential clinical relevance. Using DNA copy number alteration and RNA-seq data from head and neck squamous cell carcinoma patients from The Cancer Genome Atlas as an example, we demonstrate that MPAC predicts a patient subgroup related to immune responses not identified by analysis with either input omic data type alone. Key proteins identified via this subgroup have pathway activities related to clinical outcome as well as immune cell compositions. Our MPAC R package, available at https://bioconductor.org/packages/MPAC, enables similar multi-omic analyses on new datasets.

Single cell RNA-sequencing: A powerful yet still challenging technology to study cellular heterogeneity

Almost all biomedical research to date has relied upon mean measurements from cell populations, however it is well established that what it is observed at this macroscopic level can be the result of many interactions of several different single cells. Thus, the observable macroscopic 'average' cannot outright be used as representative of the 'average cell'. Rather, it is the resulting emerging behaviour of the actions and interactions of many different cells. Single-cell RNA sequencing (scRNA-Seq) enables the comparison of the transcriptomes of individual cells. This provides high-resolution maps of the dynamic cellular programmes allowing us to answer fundamental biological questions on their function and evolution. It also allows to address medical questions such as the role of rare cell populations contributing to disease progression and therapeutic resistance. Furthermore, it provides an understanding of context-specific dependencies, namely the behaviour and function that a cell has in a specific context, which can be crucial to understand some complex diseases, such as diabetes, cardiovascular disease and cancer. Here, we provide an overview of scRNA-Seq, including a comparative review of emerging technologies and computational pipelines. We discuss the current and emerging applications and focus on tumour heterogeneity a clear example of how scRNA-Seq can provide new understanding of a complex disease. Additionally, we review the limitations and highlight the need of powerful computational pipelines and reproducible protocols for the broader acceptance of this technique in basic and clinical research.

Undisclosed, unmet and neglected challenges in multi-omics studies

Multi-omics approaches have become a reality in both large genomics projects and small laboratories. However, the multi-omics research community still faces a number of issues that have either not been sufficiently discussed or for which current solutions are still limited. In this Perspective, we elaborate on these limitations and suggest points of attention for future research. We finally discuss new opportunities and challenges brought to the field by the rapid development of single-cell high-throughput molecular technologies.

ReactomeGSA - Efficient Multi-Omics Comparative Pathway Analysis

Pathway analyses are key methods to analyze 'omics experiments. Nevertheless, integrating data from different 'omics technologies and different species still requires considerable bioinformatics knowledge.Here we present the novel ReactomeGSA resource for comparative pathway analyses of multi-omics datasets. ReactomeGSA can be used through Reactome's existing web interface and the novel ReactomeGSA R Bioconductor package with explicit support for scRNA-seq data. Data from different species is automatically mapped to a common pathway space. Public data from ExpressionAtlas and Single Cell ExpressionAtlas can be directly integrated in the analysis. ReactomeGSA greatly reduces the technical barrier for multi-omics, cross-species, comparative pathway analyses.We used ReactomeGSA to characterize the role of B cells in anti-tumor immunity. We compared B cell rich and poor human cancer samples from five of the Cancer Genome Atlas (TCGA) transcriptomics and two of the Clinical Proteomic Tumor Analysis Consortium (CPTAC) proteomics studies. B cell-rich lung adenocarcinoma samples lacked the otherwise present activation through NFkappaB. This may be linked to the presence of a specific subset of tumor associated IgG+ plasma cells that lack NFkappaB activation in scRNA-seq data from human melanoma. This showcases how ReactomeGSA can derive novel biomedical insights by integrating large multi-omics datasets.

Comprehensive Analysis of Genetic Ancestry and Its Molecular Correlates in Cancer

We evaluated ancestry effects on mutation rates, DNA methylation, and mRNA and miRNA expression among 10,678 patients across 33 cancer types from The Cancer Genome Atlas. We demonstrated that cancer subtypes and ancestry-related technical artifacts are important confounders that have been insufficiently accounted for. Once accounted for, ancestry-associated differences spanned all molecular features and hundreds of genes. Biologically significant differences were usually tissue specific but not specific to cancer. However, admixture and pathway analyses suggested some of these differences are causally related to cancer. Specific findings included increased FBXW7 mutations in patients of African origin, decreased VHL and PBRM1 mutations in renal cancer patients of African origin, and decreased immune activity in bladder cancer patients of East Asian origin.

Uncertainty reduction in biochemical kinetic models: Enforcing desired model properties

A persistent obstacle for constructing kinetic models of metabolism is uncertainty in the kinetic properties of enzymes. Currently, available methods for building kinetic models can cope indirectly with uncertainties by integrating data from different biological levels and origins into models. In this study, we use the recently proposed computational approach iSCHRUNK (in Silico Approach to Characterization and Reduction of Uncertainty in the Kinetic Models), which combines Monte Carlo parameter sampling methods and machine learning techniques, in the context of Bayesian inference. Monte Carlo parameter sampling methods allow us to exploit synergies between different data sources and generate a population of kinetic models that are consistent with the available data and physicochemical laws. The machine learning allows us to data-mine the a priori generated kinetic parameters together with the integrated datasets and derive posterior distributions of kinetic parameters consistent with the observed physiology. In this work, we used iSCHRUNK to address a design question: can we identify which are the kinetic parameters and what are their values that give rise to a desired metabolic behavior? Such information is important for a wide variety of studies ranging from biotechnology to medicine. To illustrate the proposed methodology, we performed Metabolic Control Analysis, computed the flux control coefficients of the xylose uptake (XTR), and identified parameters that ensure a rate improvement of XTR in a glucose-xylose co-utilizing S. cerevisiae strain. Our results indicate that only three kinetic parameters need to be accurately characterized to describe the studied physiology, and ultimately to design and control the desired responses of the metabolism. This framework paves the way for a new generation of methods that will systematically integrate the wealth of available omics data and efficiently extract the information necessary for metabolic engineering and synthetic biology decisions.

Kinetic models are the most promising tool for understanding the complex dynamic behavior of living cells. The primary goal of kinetic models is to capture the properties of the metabolic networks as a whole, and thus we need large-scale models for dependable in silico analyses of metabolism. However, uncertainty in kinetic parameters impedes the development of kinetic models, and uncertainty levels increase with the model size. Tools that will address the issues with parameter uncertainty and that will be able to reduce the uncertainty propagation through the system are therefore needed. In this work, we applied a method called iSCHRUNK that combines parameter sampling and machine learning techniques to characterize the uncertainties and uncover intricate relationships between the parameters of kinetic models and the responses of the metabolic network. The proposed method allowed us to identify a small number of parameters that determine the responses in the network regardless of the values of other parameters. As a consequence, in future studies of metabolism, it will be sufficient to explore a reduced kinetic space, and more comprehensive analyses of large-scale and genome-scale metabolic networks will be computationally tractable.

Exploiting Interdata Relationships in Next-generation Proteomics Analysis

Mass spectrometry based proteomics and other technologies have matured to enable routine quantitative, system-wide analysis of concentrations, modifications, and interactions of proteins, mRNAs, and other molecules. These studies have allowed us to move toward a new field concerned with mining information from the combination of these orthogonal data sets, perhaps called "integromics." We highlight examples of recent studies and tools that aim at relating proteomic information to mRNAs, genetic associations, and changes in small molecules and lipids. We argue that productive data integration differs from parallel acquisition and interpretation and should move toward quantitative modeling of the relationships between the data. These relationships might be expressed by temporal information retrieved from time series experiments, rate equations to model synthesis and degradation, or networks of causal, evolutionary, physical, and other interactions. We outline steps and considerations toward such integromic studies to exploit the synergy between data sets.

Systems biology: perspectives on multiscale modeling in research on endocrine-related cancers

Drawing on concepts from experimental biology, computer science, informatics, mathematics and statistics, systems biologists integrate data across diverse platforms and scales of time and space to create computational and mathematical models of the integrative, holistic functions of living systems. Endocrine-related cancers are well suited to study from a systems perspective because of the signaling complexities arising from the roles of growth factors, hormones and their receptors as critical regulators of cancer cell biology and from the interactions among cancer cells, normal cells and signaling molecules in the tumor microenvironment. Moreover, growth factors, hormones and their receptors are often effective targets for therapeutic intervention, such as estrogen biosynthesis, estrogen receptors or HER2 in breast cancer and androgen receptors in prostate cancer. Given the complexity underlying the molecular control networks in these cancers, a simple, intuitive understanding of how endocrine-related cancers respond to therapeutic protocols has proved incomplete and unsatisfactory. Systems biology offers an alternative paradigm for understanding these cancers and their treatment. To correctly interpret the results of systems-based studies requires some knowledge of how in silico models are built, and how they are used to describe a system and to predict the effects of perturbations on system function. In this review, we provide a general perspective on the field of cancer systems biology, and we explore some of the advantages, limitations and pitfalls associated with using predictive multiscale modeling to study endocrine-related cancers.

Systems Biology Analyses Show Hyperactivation of Transforming Growth Factor-β and JNK Signaling Pathways in Esophageal Cancer

BACKGROUND & AIMS

Esophageal adenocarcinoma (EAC) is resistant to standard chemoradiation treatments, and few targeted therapies are available. We used large-scale tissue profiling and pharmacogenetic analyses to identify deregulated signaling pathways in EAC tissues that might be targeted to slow tumor growth or progression.

METHODS

We collected 397 biopsy specimens from patients with EAC and nonmalignant Barrett's esophagus (BE), with or without dysplasia. We performed RNA-sequencing analyses and used systems biology approaches to identify pathways that are differentially activated in EAC vs nonmalignant dysplastic tissues; pathway activities were confirmed with immunohistochemistry and quantitative real-time polymerase chain reaction analyses of signaling components in patient tissue samples. Human EAC (FLO-1 and EsoAd1), dysplastic BE (CP-B, CP-C, CP-D), and nondysplastic BE (CP-A) cells were incubated with pharmacologic inhibitors or transfected with small interfering RNAs. We measured effects on proliferation, colony formation, migration, and/or growth of xenograft tumors in nude mice.

RESULTS

Comparisons of EAC vs nondysplastic BE tissues showed hyperactivation of transforming growth factor-β (TGFB) and/or Jun N-terminal kinase (JNK) signaling pathways in more than 80% of EAC samples. Immunohistochemical analyses showed increased nuclear localization of phosphorylated JUN and SMAD proteins in EAC tumor tissues compared with nonmalignant tissues. Genes regulated by the TGFB and JNK pathway were overexpressed specifically in EAC and dysplastic BE. Pharmacologic inhibition or knockdown of TGFB or JNK signaling components in EAC cells (FLO-1 or EsoAd1) significantly reduced cell proliferation, colony formation, cell migration, and/or growth of xenograft tumors in mice in a SMAD4-independent manner. Inhibition of the TGFB pathway in BE cell lines reduced the proliferation of dysplastic, but not nondysplastic, cells.

CONCLUSIONS

In a transcriptome analysis of EAC and nondysplastic BE tissues, we found the TGFB and JNK signaling pathways to be hyperactivated in EACs and the genes regulated by these pathways to be overexpressed in EAC and dysplastic BE. Inhibiting these pathways in EAC cells reduces their proliferation, migration, and formation of xenograft tumors. Strategies to block the TGFB and JNK signaling pathways might be developed for treatment of EAC.

Computational Methods to Investigate the Impact of miRNAs on Pathways

Pathway analysis is a wide class of methods allowing to determine the alteration of functional processes in complex diseases. However, biological pathways are still partial, and knowledge coming from posttranscriptional regulators has started to be considered in a systematic way only recently. Here we will give a global and updated view of the main pathway and subpathway analysis methodologies, focusing on the improvements obtained through the recent introduction of microRNAs as regulatory elements in these frameworks.

A Comprehensive Pan-Cancer Molecular Study of Gynecologic and Breast Cancers

We analyzed molecular data on 2,579 tumors from The Cancer Genome Atlas (TCGA) of four gynecological types plus breast. Our aims were to identify shared and unique molecular features, clinically significant subtypes, and potential therapeutic targets. We found 61 somatic copy-number alterations (SCNAs) and 46 significantly mutated genes (SMGs). Eleven SCNAs and 11 SMGs had not been identified in previous TCGA studies of the individual tumor types. We found functionally significant estrogen receptor-regulated long non-coding RNAs (lncRNAs) and gene/lncRNA interaction networks. Pathway analysis identified subtypes with high leukocyte infiltration, raising potential implications for immunotherapy. Using 16 key molecular features, we identified five prognostic subtypes and developed a decision tree that classified patients into the subtypes based on just six features that are assessable in clinical laboratories.

Post-transcriptional knowledge in pathway analysis increases the accuracy of phenotypes classification

Motivation

Prediction of phenotypes from high-dimensional data is a crucial task in precision biology and medicine. Many technologies employ genomic biomarkers to characterize phenotypes. However, such elements are not sufficient to explain the underlying biology. To improve this, pathway analysis techniques have been proposed. Nevertheless, such methods have shown lack of accuracy in phenotypes classification.

Results

Here we propose a novel methodology called MITHrIL (Mirna enrIched paTHway Impact anaLysis) for the analysis of signaling pathways, which extends the work of Tarca et al., 2009. MITHrIL augments pathways with missing regulatory elements, such as microRNAs, and their interactions with genes. The method takes as input the expression values of genes and/or microRNAs and returns a list of pathways sorted according to their degree of deregulation, together with the corresponding statistical significance (p-values). Our analysis shows that MITHrIL outperforms its competitors even in the worst case. In addition, our method is able to correctly classify sets of tumor samples drawn from TCGA.

Availability

MITHrIL is freely available at the following URL: http://alpha.dmi.unict.it/mithril/

Comprehensive and Integrated Genomic Characterization of Adult Soft Tissue Sarcomas

Sarcomas are a broad family of mesenchymal malignancies exhibiting remarkable histologic diversity. We describe the multi-platform molecular landscape of 206 adult soft tissue sarcomas representing 6 major types. Along with novel insights into the biology of individual sarcoma types, we report three overarching findings: (1) unlike most epithelial malignancies, these sarcomas (excepting synovial sarcoma) are characterized predominantly by copy-number changes, with low mutational loads and only a few genes (TP53, ATRX, RB1) highly recurrently mutated across sarcoma types; (2) within sarcoma types, genomic and regulomic diversity of driver pathways defines molecular subtypes associated with patient outcome; and (3) the immune microenvironment, inferred from DNA methylation and mRNA profiles, associates with outcome and may inform clinical trials of immune checkpoint inhibitors. Overall, this large-scale analysis reveals previously unappreciated sarcoma-type-specific changes in copy number, methylation, RNA, and protein, providing insights into refining sarcoma therapy and relationships to other cancer types.

Integrative Analysis Identifies Four Molecular and Clinical Subsets in Uveal Melanoma

Comprehensive multiplatform analysis of 80 uveal melanomas (UM) identifies four molecularly distinct, clinically relevant subtypes: two associated with poor-prognosis monosomy 3 (M3) and two with better-prognosis disomy 3 (D3). We show that BAP1 loss follows M3 occurrence and correlates with a global DNA methylation state that is distinct from D3-UM. Poor-prognosis M3-UM divide into subsets with divergent genomic aberrations, transcriptional features, and clinical outcomes. We report change-of-function SRSF2 mutations. Within D3-UM, EIF1AX- and SRSF2/SF3B1-mutant tumors have distinct somatic copy number alterations and DNA methylation profiles, providing insight into the biology of these low- versus intermediate-risk clinical mutation subtypes.

minepath.org: a free interactive pathway analysis web server

Minepath

( www.minepath.org ) is a web-based platform that elaborates on, and radically extends the identification of differentially expressed sub-paths in molecular pathways. Besides the network topology, the underlying MinePath algorithmic processes exploit exact gene-gene molecular relationships (e.g. activation, inhibition) and are able to identify differentially expressed pathway parts. Each pathway is decomposed into all its constituent sub-paths, which in turn are matched with corresponding gene expression profiles. The highly ranked, and phenotype inclined sub-paths are kept. Apart from the pathway analysis algorithm, the fundamental innovation of the MinePath web-server concerns its advanced visualization and interactive capabilities. To our knowledge, this is the first pathway analysis server that introduces and offers visualization of the underlying and active pathway regulatory mechanisms instead of genes. Other features include live interaction, immediate visualization of functional sub-paths per phenotype and dynamic linked annotations for the engaged genes and molecular relations. The user can download not only the results but also the corresponding web viewer framework of the performed analysis. This feature provides the flexibility to immediately publish results without publishing source/expression data, and get all the functionality of a web based pathway analysis viewer.

Detecting Disease Specific Pathway Substructures through an Integrated Systems Biology Approach

In the era of network medicine, pathway analysis methods play a central role in the prediction of phenotype from high throughput experiments. In this paper, we present a network-based systems biology approach capable of extracting disease-perturbed subpathways within pathway networks in connection with expression data taken from The Cancer Genome Atlas (TCGA). Our system extends pathways with missing regulatory elements, such as microRNAs, and their interactions with genes. The framework enables the extraction, visualization, and analysis of statistically significant disease-specific subpathways through an easy to use web interface. Our analysis shows that the methodology is able to fill the gap in current techniques, allowing a more comprehensive analysis of the phenomena underlying disease states.

Integrated genomic and molecular characterization of cervical cancer

Cervical cancer remains one of the leading causes of cancer-related deaths worldwide. Reported here is an extensive molecular characterization of 228 primary cervical cancers, the largest comprehensive genomic study of cervical cancer to date. We observed striking APOBEC mutagenesis patterns and identified SHKBP1, ERBB3, CASP8, HLA-A, and TGFBR2 as novel significantly mutated genes in cervical cancer. We also discovered novel amplifications in immune targets CD274/PD-L1 and PDCD1LG2/PD-L2, and the BCAR4 lncRNA that has been associated with response to lapatinib. HPV integration was observed in all HPV18-related cases and 76% of HPV16-related cases, and was associated with structural aberrations and increased target gene expression. We identified a unique set of endometrial-like cervical cancers, comprised predominantly of HPV-negative tumors with high frequencies of KRAS, ARID1A, and PTEN mutations. Integrative clustering of 178 samples identified Keratin-low Squamous, Keratin-high Squamous, and Adenocarcinoma-rich subgroups. These molecular analyses reveal new potential therapeutic targets for cervical cancers.

MinePath: Mining for Phenotype Differential Sub-paths in Molecular Pathways

Pathway analysis methodologies couple traditional gene expression analysis with knowledge encoded in established molecular pathway networks, offering a promising approach towards the biological interpretation of phenotype differentiating genes. Early pathway analysis methodologies, named as gene set analysis (GSA), view pathways just as plain lists of genes without taking into account either the underlying pathway network topology or the involved gene regulatory relations. These approaches, even if they achieve computational efficiency and simplicity, consider pathways that involve the same genes as equivalent in terms of their gene enrichment characteristics. Most recent pathway analysis approaches take into account the underlying gene regulatory relations by examining their consistency with gene expression profiles and computing a score for each profile. Even with this approach, assessing and scoring single-relations limits the ability to reveal key gene regulation mechanisms hidden in longer pathway sub-paths. We introduce MinePath, a pathway analysis methodology that addresses and overcomes the aforementioned problems. MinePath facilitates the decomposition of pathways into their constituent sub-paths. Decomposition leads to the transformation of single-relations to complex regulation sub-paths. Regulation sub-paths are then matched with gene expression sample profiles in order to evaluate their functional status and to assess phenotype differential power. Assessment of differential power supports the identification of the most discriminant profiles. In addition, MinePath assess the significance of the pathways as a whole, ranking them by their p-values. Comparison results with state-of-the-art pathway analysis systems are indicative for the soundness and reliability of the MinePath approach. In contrast with many pathway analysis tools, MinePath is a web-based system (www.minepath.org) offering dynamic and rich pathway visualization functionality, with the unique characteristic to color regulatory relations between genes and reveal their phenotype inclination. This unique characteristic makes MinePath a valuable tool for in silico molecular biology experimentation as it serves the biomedical researchers' exploratory needs to reveal and interpret the regulatory mechanisms that underlie and putatively govern the expression of target phenotypes.

Learning mixed graphical models with separate sparsity parameters and stability-based model selection

BACKGROUND

Mixed graphical models (MGMs) are graphical models learned over a combination of continuous and discrete variables. Mixed variable types are common in biomedical datasets. MGMs consist of a parameterized joint probability density, which implies a network structure over these heterogeneous variables. The network structure reveals direct associations between the variables and the joint probability density allows one to ask arbitrary probabilistic questions on the data. This information can be used for feature selection, classification and other important tasks.

RESULTS

We studied the properties of MGM learning and applications of MGMs to high-dimensional data (biological and simulated). Our results show that MGMs reliably uncover the underlying graph structure, and when used for classification, their performance is comparable to popular discriminative methods (lasso regression and support vector machines). We also show that imposing separate sparsity penalties for edges connecting different types of variables significantly improves edge recovery performance. To choose these sparsity parameters, we propose a new efficient model selection method, named Stable Edge-specific Penalty Selection (StEPS). StEPS is an expansion of an earlier method, StARS, to mixed variable types. In terms of edge recovery, StEPS selected MGMs outperform those models selected using standard techniques, including AIC, BIC and cross-validation. In addition, we use a heuristic search that is linear in size of the sparsity value search space as opposed to the cubic grid search required by other model selection methods. We applied our method to clinical and mRNA expression data from the Lung Genomics Research Consortium (LGRC) and the learned MGM correctly recovered connections between the diagnosis of obstructive or interstitial lung disease, two diagnostic breathing tests, and cigarette smoking history. Our model also suggested biologically relevant mRNA markers that are linked to these three clinical variables.

CONCLUSIONS

MGMs are able to accurately recover dependencies between sets of continuous and discrete variables in both simulated and biomedical datasets. Separation of sparsity penalties by edge type is essential for accurate network edge recovery. Furthermore, our stability based method for model selection determines sparsity parameters faster and more accurately (in terms of edge recovery) than other model selection methods. With the ongoing availability of comprehensive clinical and biomedical datasets, MGMs are expected to become a valuable tool for investigating disease mechanisms and answering an array of critical healthcare questions.

Network-based methods for identifying critical pathways of complex diseases: a survey

We review seven major network-based pathway analysis methods and enumerate their benefits and limitations from an algorithmic perspective to provide a reference for the next generation of pathway analysis methods.

DGCR8 is essential for tumor progression following PTEN loss in the prostate

In human prostate cancer, the microRNA biogenesis machinery increases with prostate cancer progression. Here, we show that deletion of the Dgcr8 gene, a critical component of this complex, inhibits tumor progression in a Pten-knockout mouse model of prostate cancer. Early stages of tumor development were unaffected, but progression to advanced prostatic intraepithelial neoplasia was severely inhibited. Dgcr8 loss blocked Pten null-induced expansion of the basal-like, but not luminal, cellular compartment. Furthermore, while late-stage Pten knockout tumors exhibit decreased senescence-associated beta-galactosidase activity and increased proliferation, the simultaneous deletion of Dgcr8 blocked these changes resulting in levels similar to wild type. Sequencing of small RNAs in isolated epithelial cells uncovered numerous miRNA changes associated with PTEN loss. Consistent with a Pten-Dgcr8 association, analysis of a large cohort of human prostate tumors shows a strong correlation between Akt activation and increased Dgcr8 mRNA levels. Together, these findings uncover a critical role for microRNAs in enhancing proliferation and enabling the expansion of the basal cell compartment associated with tumor progression following Pten loss.

Sharing information to reconstruct patient-specific pathways in heterogeneous diseases

Advances in experimental techniques resulted in abundant genomic, transcriptomic, epigenomic, and proteomic data that have the potential to reveal critical drivers of human diseases. Complementary algorithmic developments enable researchers to map these data onto protein-protein interaction networks and infer which signaling pathways are perturbed by a disease. Despite this progress, integrating data across different biological samples or patients remains a substantial challenge because samples from the same disease can be extremely heterogeneous. Somatic mutations in cancer are an infamous example of this heterogeneity. Although the same signaling pathways may be disrupted in a cancer patient cohort, the distribution of mutations is long-tailed, and many driver mutations may only be detected in a small fraction of patients. We developed a computational approach to account for heterogeneous data when inferring signaling pathways by sharing information across the samples. Our technique builds upon the prize-collecting Steiner forest problem, a network optimization algorithm that extracts pathways from a protein-protein interaction network. We recover signaling pathways that are similar across all samples yet still reflect the unique characteristics of each biological sample. Leveraging data from related tumors improves our ability to recover the disrupted pathways and reveals patient-specific pathway perturbations in breast cancer.