Pathway-based classification of cancer subtypes

Big data to smart data in Alzheimer's disease: Real-world examples of advanced modeling and simulation

Many disease-modifying clinical development programs in Alzheimer's disease (AD) have failed to date, and development of new and advanced preclinical models that generate actionable knowledge is desperately needed. This review reports on computer-based modeling and simulation approach as a powerful tool in AD research. Statistical data-analysis techniques can identify associations between certain data and phenotypes, such as diagnosis or disease progression. Other approaches integrate domain expertise in a formalized mathematical way to understand how specific components of pathology integrate into complex brain networks. Private-public partnerships focused on data sharing, causal inference and pathway-based analysis, crowdsourcing, and mechanism-based quantitative systems modeling represent successful real-world modeling examples with substantial impact on CNS diseases. Similar to other disease indications, successful real-world examples of advanced simulation can generate actionable support of drug discovery and development in AD, illustrating the value that can be generated for different stakeholders.

Reconstruction of Tissue-Specific Metabolic Networks Using CORDA

Human metabolism involves thousands of reactions and metabolites. To interpret this complexity, computational modeling becomes an essential experimental tool. One of the most popular techniques to study human metabolism as a whole is genome scale modeling. A key challenge to applying genome scale modeling is identifying critical metabolic reactions across diverse human tissues. Here we introduce a novel algorithm called Cost Optimization Reaction Dependency Assessment (CORDA) to build genome scale models in a tissue-specific manner. CORDA performs more efficiently computationally, shows better agreement to experimental data, and displays better model functionality and capacity when compared to previous algorithms. CORDA also returns reaction associations that can greatly assist in any manual curation to be performed following the automated reconstruction process. Using CORDA, we developed a library of 76 healthy and 20 cancer tissue-specific reconstructions. These reconstructions identified which metabolic pathways are shared across diverse human tissues. Moreover, we identified changes in reactions and pathways that are differentially included and present different capacity profiles in cancer compared to healthy tissues, including up-regulation of folate metabolism, the down-regulation of thiamine metabolism, and tight regulation of oxidative phosphorylation.

An Optimization-Based Framework for the Transformation of Incomplete Biological Knowledge into a Probabilistic Structure and Its Application to the Utilization of Gene/Protein Signaling Pathways in Discrete Phenotype Classification

Phenotype classification via genomic data is hampered by small sample sizes that negatively impact classifier design. Utilization of prior biological knowledge in conjunction with training data can improve both classifier design and error estimation via the construction of the optimal Bayesian classifier. In the genomic setting, gene/protein signaling pathways provide a key source of biological knowledge. Although these pathways are neither complete, nor regulatory, with no timing associated with them, they are capable of constraining the set of possible models representing the underlying interaction between molecules. The aim of this paper is to provide a framework and the mathematical tools to transform signaling pathways to prior probabilities governing uncertainty classes of feature-label distributions used in classifier design. Structural motifs extracted from the signaling pathways are mapped to a set of constraints on a prior probability on a Multinomial distribution. Being the conjugate prior for the Multinomial distribution, we propose optimization paradigms to estimate the parameters of a Dirichlet distribution in the Bayesian setting. The performance of the proposed methods is tested on two widely studied pathways: mammalian cell cycle and a p53 pathway model.

Highly informative marker sets consisting of genes with low individual degree of differential expression

Genes with significant differential expression are traditionally used to reveal the genetic background underlying phenotypic differences between cancer cells. We hypothesized that informative marker sets can be obtained by combining genes with a relatively low degree of individual differential expression. We developed a method for construction of highly informative gene combinations aimed at the maximization of the cumulative informative power and identified sets of 2–5 genes efficiently predicting recurrence for ER-positive breast cancer patients. The gene combinations constructed on the basis of microarray data were successfully applied to data acquired by RNA-seq. The developed method provides the basis for the generation of highly efficient prognostic and predictive gene signatures for cancer and other diseases. The identified gene sets can potentially reveal novel essential segments of gene interaction networks and pathways implied in cancer progression.

Deregulation of F-box proteins and its consequence on cancer development, progression and metastasis

F-box proteins are substrate receptors of the SCF (SKP1-Cullin 1-F-box protein) E3 ubiquitin ligase that play important roles in a number of physiological processes and activities. Through their ability to assemble distinct E3 ubiquitin ligases and target key regulators of cellular activities for ubiquitylation and degradation, this versatile group of proteins is able to regulate the abundance of cellular proteins whose deregulated expression or activity contributes to disease. In this review, we describe the important roles of select F-box proteins in regulating cellular activities, the perturbation of which contributes to the initiation and progression of a number of human malignancies.

Topologically inferring pathway activity toward precise cancer classification via integrating genomic and metabolomic data: prostate cancer as a case

Precise cancer classification is a central challenge in clinical cancer research such as diagnosis, prognosis and metastasis prediction. Most existing cancer classification methods based on gene or metabolite biomarkers were limited to single genomics or metabolomics, and lacked integration and utilization of multiple ‘omics’ data. The accuracy and robustness of these methods when applied to independent cohorts of patients must be improved. In this study, we propose a directed random walk-based method to evaluate the topological importance of each gene in a reconstructed gene–metabolite graph by integrating information from matched gene expression profiles and metabolomic profiles. The joint use of gene and metabolite information contributes to accurate evaluation of the topological importance of genes and reproducible pathway activities. We constructed classifiers using reproducible pathway activities for precise cancer classification and risk metabolic pathway identification. We applied the proposed method to the classification of prostate cancer. Within-dataset experiments and cross-dataset experiments on three independent datasets demonstrated that the proposed method achieved a more accurate and robust overall performance compared to several existing classification methods. The resulting risk pathways and topologically important differential genes and metabolites provide biologically informative models for prostate cancer prognosis and therapeutic strategies development.

Machine learning for biomarker identification in cancer research - developments toward its clinical application

The patterns identified from the systematically collected molecular profiles of patient tumor samples, along with clinical metadata, can assist personalized treatments for effective management of cancer patients with similar molecular subtypes. There is an unmet need to develop computational algorithms for cancer diagnosis, prognosis and therapeutics that can identify complex patterns and help in classifications based on plethora of emerging cancer research outcomes in public domain. Machine learning, a branch of artificial intelligence, holds a great potential for pattern recognition in cryptic cancer datasets, as evident from recent literature survey. In this review, we focus on the current status of machine learning applications in cancer research, highlighting trends and analyzing major achievements, roadblocks and challenges toward its implementation in clinics.

Using prior knowledge from cellular pathways and molecular networks for diagnostic specimen classification

For many complex diseases, an earlier and more reliable diagnosis is considered a key prerequisite for developing more effective therapies to prevent or delay disease progression. Classical statistical learning approaches for specimen classification using omics data, however, often cannot provide diagnostic models with sufficient accuracy and robustness for heterogeneous diseases like cancers or neurodegenerative disorders. In recent years, new approaches for building multivariate biomarker models on omics data have been proposed, which exploit prior biological knowledge from molecular networks and cellular pathways to address these limitations. This survey provides an overview of these recent developments and compares pathway- and network-based specimen classification approaches in terms of their utility for improving model robustness, accuracy and biological interpretability. Different routes to translate omics-based multifactorial biomarker models into clinical diagnostic tests are discussed, and a previous study is presented as example.

Classification by integrating plant stress response gene expression data with biological knowledge

Classification of microarray data has always been a challenging task because of the enormous number of genes. In this study, a clustering method by integrating plant stress response gene expression data with biological knowledge is presented. Clustering is one of the promising tools for attribute reduction, but gene clusters are biologically uninformative. So we integrated biological knowledge into genomic analysis to help to improve the interpretation of the results. Biological similarity based on gene ontology (GO) semantic similarity was combined with gene expression data to find out biologically meaningful clusters. Affinity propagation clustering algorithm was chosen to analyze the impact of the biological similarity on the results. Based on clustering result, neighborhood rough set was used to select representative genes for each cluster. The prediction accuracy of classifiers built on reduced gene subsets indicated that our approach outperformed other classical methods. The information fusion was proven to be effective through quantitative analysis, as it could select gene subsets with high biological significance and select significant genes.

Multiclass cancer classification based on gene expression comparison

As the complexity and heterogeneity of cancer is being increasingly appreciated through genomic analyses, microarray-based cancer classification comprising multiple discriminatory molecular markers is an emerging trend. Such multiclass classification problems pose new methodological and computational challenges for developing novel and effective statistical approaches. In this paper, we introduce a new approach for classifying multiple disease states associated with cancer based on gene expression profiles. Our method focuses on detecting small sets of genes in which the relative comparison of their expression values leads to class discrimination. For an m-class problem, the classification rule typically depends on a small number of m-gene sets, which provide transparent decision boundaries and allow for potential biological interpretations. We first test our approach on seven common gene expression datasets and compare it with popular classification methods including support vector machines and random forests. We then consider an extremely large cohort of leukemia cancer patients to further assess its effectiveness. In both experiments, our method yields comparable or even better results to benchmark classifiers. In addition, we demonstrate that our approach can integrate pathway analysis of gene expression to provide accurate and biological meaningful classification.

Identification of Novel Breast Cancer Subtype-Specific Biomarkers by Integrating Genomics Analysis of DNA Copy Number Aberrations and miRNA-mRNA Dual Expression Profiling

Breast cancer is a heterogeneous disease with well-defined molecular subtypes. Currently, comparative genomic hybridization arrays (aCGH) techniques have been developed rapidly, and recent evidences in studies of breast cancer suggest that tumors within gene expression subtypes share similar DNA copy number aberrations (CNA) which can be used to further subdivide subtypes. Moreover, subtype-specific miRNA expression profiles are also proposed as novel signatures for breast cancer classification. The identification of mRNA or miRNA expression-based breast cancer subtypes is considered an instructive means of prognosis. Here, we conducted an integrated analysis based on copy number aberrations data and miRNA-mRNA dual expression profiling data to identify breast cancer subtype-specific biomarkers. Interestingly, we found a group of genes residing in subtype-specific CNA regions that also display the corresponding changes in mRNAs levels and their target miRNAs' expression. Among them, the predicted direct correlation of BRCA1-miR-143-miR-145 pairs was selected for experimental validation. The study results indicated that BRCA1 positively regulates miR-143-miR-145 expression and miR-143-miR-145 can serve as promising novel biomarkers for breast cancer subtyping. In our integrated genomics analysis and experimental validation, a new frame to predict candidate biomarkers of breast cancer subtype is provided and offers assistance in order to understand the potential disease etiology of the breast cancer subtypes.

Inhibitor of differentiation 4 (ID4) acts as an inhibitor of ID-1, -2 and -3 and promotes basic helix loop helix (bHLH) E47 DNA binding and transcriptional activity

The four known ID proteins (ID1-4, Inhibitor of Differentiation) share a homologous helix loop helix (HLH) domain and act as dominant negative regulators of basic-HLH transcription factors. ID proteins also interact with many non-bHLH proteins in complex networks. The expression of ID proteins is increasingly observed in many cancers. Whereas ID-1, ID-2 and ID-3, are generally considered as tumor promoters, ID4 on the contrary has emerged as a tumor suppressor. In this study we demonstrate that ID4 heterodimerizes with ID-1, -2 and -3 and promote bHLH DNA binding, essentially acting as an inhibitor of inhibitors of differentiation proteins. Interaction of ID4 was observed with ID1, ID2 and ID3 that was dependent on intact HLH domain of ID4. Interaction with bHLH protein E47 required almost 3 fold higher concentration of ID4 as compared to ID1. Furthermore, inhibition of E47 DNA binding by ID1 was restored by ID4 in an EMSA binding assay. ID4 and ID1 were also colocalized in prostate cancer cell line LNCaP. The alpha helix forming alanine stretch N-terminal, unique to HLH ID4 domain was required for optimum interaction. Ectopic expression of ID4 in DU145 prostate cancer line promoted E47 dependent expression of CDKNI p21. Thus counteracting the biological activities of ID-1, -2 and -3 by forming inactive heterodimers appears to be a novel mechanism of action of ID4. These results could have far reaching consequences in developing strategies to target ID proteins for cancer therapy and understanding biologically relevant ID-interactions.

F-box protein Fbxl18 mediates polyubiquitylation and proteasomal degradation of the pro-apoptotic SCF subunit Fbxl7

Fbxl7, a subunit of the SCF (Skp-Cul1-F-box protein) complex induces mitotic arrest in cells; however, molecular factors that control its cellular abundance remain largely unknown. Here, we identified that an orphan F-box protein, Fbxl18, targets Fbxl7 for its polyubiquitylation and proteasomal degradation. Lys 109 within Fbxl7 is an essential acceptor site for ubiquitin conjugation by Fbxl18. An FQ motif within Fbxl7 serves as a molecular recognition site for Fbxl18 interaction. Ectopically expressed Fbxl7 induces apoptosis in Hela cells, an effect profoundly accentuated after cellular depletion of Fbxl18 protein or expression of Fbxl7 plasmids encoding mutations at either Lys 109 or within the FQ motif. Ectopic expression of Fbxl18 plasmid-limited apoptosis caused by overexpressed Fbxl7 plasmid. Thus, Fbxl18 regulates apoptosis by mediating ubiquitin-dependent proteasomal degradation of the pro-apoptotic protein Fbxl7 that may impact cellular processes involved in cell cycle progression.

Inhibitor of differentiation 4 (ID4): From development to cancer

Highly conserved Inhibitors of DNA-Binding (ID1-ID4) genes encode multi-functional proteins whose transcriptional activity is based on dominant negative inhibition of basic helix-loop-helix (bHLH) transcription factors. Initial animal models indicated a degree of compensatory overlap between ID genes such that deletion of multiple ID genes was required to generate easily recognizable phenotypes. More recently, new model systems have revealed alterations in mice harboring deletions in single ID genes suggesting complex gene and tissue specific functions for members of the ID gene family. Because ID genes are highly expressed during development and their function is associated with a primitive, proliferative cellular phenotype there has been significant interest in understanding their potential roles in neoplasia. Indeed, numerous studies indicate an oncogenic function for ID1, ID2 and ID3. In contrast, the inhibitor of differentiation 4 (ID4) presents a paradigm shift in context of well-established role of ID1, ID2 and ID3 in development and cancer. Apart from some degree of functional redundancy such as HLH dependent interactions with bHLH protein E2A, many of the functions of ID4 are distinct from ID1, ID2 and ID3: ID4 proteins a) regulate distinct developmental processes and tissue expression in the adult, b) promote stem cell survival, differentiation and/or timing of differentiation, c) epigenetic inactivation/loss of expression in several advanced stage cancers and d) increased expression in some cancers such as those arising in the breast and ovary. Thus, in spite of sharing the conserved HLH domain, ID4 defies the established model of ID protein function and expression. The underlying molecular mechanism responsible for the unique role of ID4 as compared to other ID proteins still remains largely un-explored. This review will focus on the current understanding of ID4 in context of development and cancer.

MicroRNA signatures highlight new breast cancer subtypes

MicroRNAs (miRNAs) are a kind of short non-coding RNAs, of about 22 nucleotides in length, which modulate and sometimes degrade the target mRNAs thereby regulating a number of cellular functions. Recent research in this area establishes the involvement of miRNAs in various disease progressions, including certain types of cancer development. Further, genome-wide expression profiling of miRNAs has been proven to be useful for differentiating various cancer types. In this paper, we have used miRNA expression profiles over a large set of breast cancer tumor samples for identifying subtypes of breast cancers. The experimental results demonstrate that miRNAs carry a unique signature that distinguishes cancer subtypes and reveal new cancer subtypes. Additional survival analyses based on clinical data also strengthen this claim.

Recent Advancements in Prognostic Factors of Epithelial Ovarian Carcinoma

Ovarian cancer remains the most common cause of gynecologic cancer-related death among women in developed countries. Nevertheless, subgroups of ovarian cancer patients experience relatively longer survival. Efforts to identify prognostic factors that characterize such patients are ongoing, with investigational areas including tumor characteristics, surgical management, inheritance patterns, immunologic factors, and genomic patterns. This review discusses various demographic, clinical, and molecular factors implicating longevity and ovarian cancer survival. Continued efforts at identifying these prognosticators may result in invaluable adjuncts to the treatment of ovarian cancer, with the ultimate goal of advancing patient care.

Integrative Pathway Analysis Using Graph-Based Learning with Applications to TCGA Colon and Ovarian Data

Recent method development has included multi-dimensional genomic data algorithms because such methods have more accurately predicted clinical phenotypes related to disease. This study is the first to conduct an integrative genomic pathway-based analysis with a graph-based learning algorithm. The methodology of this analysis, graph-based semi-supervised learning, detects pathways that improve prediction of a dichotomous variable, which in this study is cancer stage. This analysis integrates genome-level gene expression, methylation, and single nucleotide polymorphism (SNP) data in serous cystadenocarcinoma (OV) and colon adenocarcinoma (COAD). The top 10 ranked predictive pathways in COAD and OV were biologically relevant to their respective cancer stages and significantly enhanced prediction accuracy and area under the ROC curve (AUC) when compared to single data-type analyses. This method is an effective way to simultaneously predict binary clinical phenotypes and discover their biological mechanisms.

Functional characterization of breast cancer using pathway profiles

BACKGROUND

The molecular characteristics of human diseases are often represented by a list of genes termed "signature genes". A significant challenge facing this approach is that of reproducibility: signatures developed on a set of patients may fail to perform well on different sets of patients. As diseases are resulted from perturbed cellular functions, irrespective of the particular genes that contribute to the function, it may be more appropriate to characterize diseases based on these perturbed cellular functions.

METHODS

We proposed a profile-based approach to characterize a disease using a binary vector whose elements indicate whether a given function is perturbed based on the enrichment analysis of expression data between normal and tumor tissues. Using breast cancer and its four primary clinically relevant subtypes as examples, this approach is evaluated based on the reproducibility, accuracy and resolution of the resulting pathway profiles.

RESULTS

Pathway profiles for breast cancer and its subtypes are constructed based on data obtained from microarray and RNA-Seq data sets provided by The Cancer Genome Atlas (TCGA), and an additional microarray data set provided by The European Genome-phenome Archive (EGA). An average reproducibility of 68% is achieved between different data sets (TCGA microarray vs. EGA microarray data) and 67% average reproducibility is achieved between different technologies (TCGA microarray vs. TCGA RNA-Seq data). Among the enriched pathways, 74% of them are known to be associated with breast cancer or other cancers. About 40% of the identified pathways are enriched in all four subtypes, with 4, 2, 4, and 7 pathways enriched only in luminal A, luminal B, triple-negative, and HER2+ subtypes, respectively. Comparison of profiles between subtypes, as well as other diseases, shows that luminal A and luminal B subtypes are more similar to the HER2+ subtype than to the triple-negative subtype, and subtypes of breast cancer are more likely to be closer to each other than to other diseases.

CONCLUSIONS

Our results demonstrate that pathway profiles can successfully characterize both common and distinct functional characteristics of four subtypes of breast cancer and other related diseases, with acceptable reproducibility, high accuracy and reasonable resolution.

Expression and significance of squalene epoxidase in squamous lung cancerous tissues and pericarcinoma tissues

BACKGROUND

A high expression of squalene epoxidase (SQLE) is related to tumor occurrence, development, and prognosis in a variety of cancers. In this study, the expression and significance of SQLE was analyzed in patients with squamous lung cancer and pericarcinoma tissues.

METHODS

The SQLE mRNA and protein expression were separately examined by semi-quantitative reverse transcription polymerase chain reaction, fluorescence quantitative polymerase chain reaction and Western blot in 65 cases of squamous cell lung carcinoma tissues and adjacent non-cancerous lung tissues.

RESULTS

The expression of SQLE mRNA and protein in lung squamous cancerous tissues was significantly higher than in pericarcinoma tissues (63.07% vs. 44.61%, P = 0.0348; 67.69% vs. 38.46%, P = 0.0008). The positive expression rate of SQLE mRNA was not associated with gender, age, smoking, or tumor size (P > 0.05). The expression of SQLE mRNA was closely correlated with poor differentiation, clinical stages, and lymphatic metastasis (P < 0.05). The expression of SQLE mRNA was negatively associated with overall survival rate (P < 0.05).

CONCLUSION

A high expression of SQLE was found in squamous lung cancer tissues. The high expression of the SQLE gene may be closely related to the occurrence and development of squamous cell lung carcinoma. SQLE expression predicts a poor prognosis and may serve as a novel lung carcinoma molecule marker.

Skin sensitizers differentially regulate signaling pathways in MUTZ-3 cells in relation to their individual potency

Background

Due to the recent European legislations posing a ban of animal tests for safety assessment within the cosmetic industry, development of in vitro alternatives for assessment of skin sensitization is highly prioritized. To date, proposed in vitro assays are mainly based on single biomarkers, which so far have not been able to classify and stratify chemicals into subgroups, related to risk or potency.

Methods

Recently, we presented the Genomic Allergen Rapid Detection (GARD) assay for assessment of chemical sensitizers. In this paper, we show how the genome wide readout of GARD can be expanded and used to identify differentially regulated pathways relating to individual chemical sensitizers. In this study, we investigated the mechanisms of action of a range of skin sensitizers through pathway identification, pathway classification and transcription factor analysis and related this to the reactive mechanisms and potency of the sensitizing agents.

Results

By transcriptional profiling of chemically stimulated MUTZ-3 cells, 33 canonical pathways intimately involved in sensitization to chemical substances were identified. The results showed that metabolic processes, cell cycling and oxidative stress responses are the key events activated during skin sensitization, and that these functions are engaged differently depending on the reactivity mechanisms of the sensitizing agent. Furthermore, the results indicate that the chemical reactivity groups seem to gradually engage more pathways and more molecules in each pathway with increasing sensitizing potency of the chemical used for stimulation. Also, a switch in gene regulation from up to down regulation, with increasing potency, was seen both in genes involved in metabolic functions and cell cycling. These observed pathway patterns were clearly reflected in the regulatory elements identified to drive these processes, where 33 regulatory elements have been proposed for further analysis.

Conclusions

This study demonstrates that functional analysis of biomarkers identified from our genomics study of human MUTZ-3 cells can be used to assess sensitizing potency of chemicals in vitro, by the identification of key cellular events, such as metabolic and cell cycling pathways.

Incorporation of Biological Pathway Knowledge in the Construction of Priors for Optimal Bayesian Classification

Small samples are commonplace in genomic/proteomic classification, the result being inadequate classifier design and poor error estimation. The problem has recently been addressed by utilizing prior knowledge in the form of a prior distribution on an uncertainty class of feature-label distributions. A critical issue remains: how to incorporate biological knowledge into the prior distribution. For genomics/proteomics, the most common kind of knowledge is in the form of signaling pathways. Thus, it behooves us to find methods of transforming pathway knowledge into knowledge of the feature-label distribution governing the classification problem. In this paper, we address the problem of prior probability construction by proposing a series of optimization paradigms that utilize the incomplete prior information contained in pathways (both topological and regulatory). The optimization paradigms employ the marginal log-likelihood, established using a small number of feature-label realizations (sample points) regularized with the prior pathway information about the variables. In the special case of a Normal-Wishart prior distribution on the mean and inverse covariance matrix (precision matrix) of a Gaussian distribution, these optimization problems become convex. Companion website: gsp.tamu.edu/Publications/supplementary/shahrokh13a.

Pathway-based outlier method reveals heterogeneous genomic structure of autism in blood transcriptome

Background

Decades of research strongly suggest that the genetic etiology of autism spectrum disorders (ASDs) is heterogeneous. However, most published studies focus on group differences between cases and controls. In contrast, we hypothesized that the heterogeneity of the disorder could be characterized by identifying pathways for which individuals are outliers rather than pathways representative of shared group differences of the ASD diagnosis.

Methods

Two previously published blood gene expression data sets – the Translational Genetics Research Institute (TGen) dataset (70 cases and 60 unrelated controls) and the Simons Simplex Consortium (Simons) dataset (221 probands and 191 unaffected family members) – were analyzed. All individuals of each dataset were projected to biological pathways, and each sample’s Mahalanobis distance from a pooled centroid was calculated to compare the number of case and control outliers for each pathway.

Results

Analysis of a set of blood gene expression profiles from 70 ASD and 60 unrelated controls revealed three pathways whose outliers were significantly overrepresented in the ASD cases: neuron development including axonogenesis and neurite development (29% of ASD, 3% of control), nitric oxide signaling (29%, 3%), and skeletal development (27%, 3%). Overall, 50% of cases and 8% of controls were outliers in one of these three pathways, which could not be identified using group comparison or gene-level outlier methods. In an independently collected data set consisting of 221 ASD and 191 unaffected family members, outliers in the neurogenesis pathway were heavily biased towards cases (20.8% of ASD, 12.0% of control). Interestingly, neurogenesis outliers were more common among unaffected family members (Simons) than unrelated controls (TGen), but the statistical significance of this effect was marginal (Chi squared P < 0.09).

Conclusions

Unlike group difference approaches, our analysis identified the samples within the case and control groups that manifested each expression signal, and showed that outlier groups were distinct for each implicated pathway. Moreover, our results suggest that by seeking heterogeneity, pathway-based outlier analysis can reveal expression signals that are not apparent when considering only shared group differences.

Topologically inferring risk-active pathways toward precise cancer classification by directed random walk

MOTIVATION

The accurate prediction of disease status is a central challenge in clinical cancer research. Microarray-based gene biomarkers have been identified to predict outcome and outperform traditional clinical parameters. However, the robustness of the individual gene biomarkers is questioned because of their little reproducibility between different cohorts of patients. Substantial progress in treatment requires advances in methods to identify robust biomarkers. Several methods incorporating pathway information have been proposed to identify robust pathway markers and build classifiers at the level of functional categories rather than of individual genes. However, current methods consider the pathways as simple gene sets but ignore the pathway topological information, which is essential to infer a more robust pathway activity.

RESULTS

Here, we propose a directed random walk (DRW)-based method to infer the pathway activity. DRW evaluates the topological importance of each gene by capturing the structure information embedded in the directed pathway network. The strategy of weighting genes by their topological importance greatly improved the reproducibility of pathway activities. Experiments on 18 cancer datasets showed that the proposed method yielded a more accurate and robust overall performance compared with several existing gene-based and pathway-based classification methods. The resulting risk-active pathways are more reliable in guiding therapeutic selection and the development of pathway-specific therapeutic strategies.

AVAILABILITY

DRW is freely available at http://210.46.85.180:8080/DRWPClass/

Quantitative proteomic analysis identifies CPNE3 as a novel metastasis-promoting gene in NSCLC

To discover metastasis-associated proteins within cancer cells, we used the isobaric tags for relative and absolute quantitation (iTRAQ) approach combined with nano liquid chromatography-tandem mass spectrometry (NanoLC-MS/MS) analysis to identify proteins that were differentially expressed between lung adenocarcinoma cancer cell lines SPC-A-1sci cells with high metastatic potential and parent SPC-A-1 cells with low metastatic potential. By employing biological and technical replicates, we identified 5818 nonredundant proteins and quantified 5443 proteins, 256 of which were differentially expressed in the two cell lines. Through si-RNA-mediated functional screens, Myosin heavy chain 9 (MYH9) and Copine III (CPNE3) were indicated as positively correlating with the migration and invasion properties of SPC-A1sci cells, and the same function of CPNE3 was confirmed in another lung cancer cell line, H1299. Furthermore, overexpressing CPNE3 promoted nonsmall-cell lung cancer (NSCLC) cell line (SPC-A-1 and XL-2) migration and invasion in vitro. Moreover, the targeted knock-down of CPNE3 inhibited the in vivo metastatic abilities of H1299 cells in mouse models. Lastly, immunohistochemistry revealed that the CPNE3 expression level was positively correlated with the clinical stage and TNM classification in NSCLC patients. Taken together, our results indicate that CPNE3 could play a critical role in NSCLC metastasis.