Extracting consistent knowledge from highly inconsistent cancer gene data sources

In silico Prediction on the PI3K/AKT/mTOR Pathway of the Antiproliferative Effect of O. joconostle in Breast Cancer Models

The search for new cancer treatments from traditional medicine involves developing studies to understand at the molecular level different cell signaling pathways involved in cancer development. In this work, we present a model of the PI3K/Akt/mTOR pathway, which plays a key role in cell cycle regulation and is related to cell survival, proliferation, and growth in cancer, as well as resistance to antitumor therapies, so finding drugs that act on this pathway is ideal to propose a new adjuvant treatment. The aim of this work was to model, simulate and predict in silico using the Big Data-Cellulat platform the possible targets in the PI3K/Akt/mTOR pathway on which the Opuntia joconostle extract acts, as well as to indicate the concentration range to be used to find the mean lethal dose in in vitro experiments on breast cancer cells. The in silico results show that, in a cancer cell, the activation of JAK and STAT, as well as PI3K and Akt is related to the effect of cell proliferation, angiogenesis, and inhibition of apoptosis, and that the extract of O. joconostle has an antiproliferative effect on breast cancer cells by inhibiting cell proliferation, regulating the cell cycle and inhibiting apoptosis through this signaling pathway . In vitro it was demonstrated that the extract shows an antiproliferative effect, causing the arrest of cells in the G2/M phase of the cell cycle. Therefore, it is concluded that the use of in silico tools is a valuable method to perform virtual experiments and discover new treatments. The use of this type of model supports in vitro experimentation, reducing the costs and number of experiments in the real laboratory.

Identification of Genes Universally Differentially Expressed in Gastric Cancer

Owing to the remarkable heterogeneity of gastric cancer (GC), population-level differentially expressed genes (DEGs) identified using case-control comparison cannot indicate the dysregulated frequency of each DEG in GC. In this work, first, the individual-level DEGs were identified for 1,090 GC tissues without paired normal tissues using the RankComp method. Second, we directly compared the gene expression in a cancer tissue to that in paired normal tissue to identify individual-level DEGs among 448 paired cancer-normal gastric tissues. We found 25 DEGs to be dysregulated in more than 90% of 1,090 GC tissues and also in more than 90% of 448 GC tissues with paired normal tissues. The 25 genes were defined as universal DEGs for GC. Then, we measured 24 paired cancer-normal gastric tissues by RNA-seq to validate them further. Among the universal DEGs, 4 upregulated genes (BGN, E2F3, PLAU, and SPP1) and 1 downregulated gene (UBL3) were found to be cancer genes already documented in the COSMIC or F-Census databases. By analyzing protein-protein interaction networks, we found 12 universally upregulated genes, and we found that their 284 direct neighbor genes were significantly enriched with cancer genes and key biological pathways related to cancer, such as the MAPK signaling pathway, cell cycle, and focal adhesion. The 13 universally downregulated genes and 16 direct neighbor genes were also significantly enriched with cancer genes and pathways related to gastric acid secretion. These universal DEGs may be of special importance to GC diagnosis and treatment targets, and they may make it easier to study the molecular mechanisms underlying GC.

contamDE-lm: linear model-based differential gene expression analysis using next-generation RNA-seq data from contaminated tumor samples

MOTIVATION

Tumor and adjacent normal RNA samples are commonly used to screen differentially expressed genes between normal and tumor samples or among tumor subtypes. Such paired-sample design could avoid numerous confounders in differential expression (DE) analysis, but the cellular contamination of tumor samples can be an important noise and confounding factor, which can both inflate false-positive rate and deflate true-positive rate. The existing DE tools that use next-generation RNA-seq data either do not account for cellular contamination or are computationally extensive with increasingly large sample size.

RESULTS

A novel linear model was proposed to avoid the problem that could arise from tumor-normal correlation for paired samples. A statistically robust and computationally very fast DE analysis procedure, contamDE-lm, was developed based on the novel model to account for cellular contamination, boosting DE analysis power through the reduction in individual residual variances using gene-wise information. The desired advantages of contamDE-lm over some state-of-the-art methods (limma and DESeq2) were evaluated through the applications to simulated data, TCGA database and Gene Expression Omnibus (GEO) database.

AVAILABILITY AND IMPLEMENTATION

The proposed method contamDE-lm was implemented in an updated R package contamDE (version 2.0), which is freely available at https://github.com/zhanghfd/contamDE.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Protein Domain Level Cancer Drug Targets in the Network of MAPK Pathways

Proteins in the MAPK pathways considered as potential drug targets for cancer treatment. Pathways along with the cross-talks increase their scope to view them as a network of MAPK pathways. Side effect causing targeted domains act as a proxy for drug targets due to its structural similarity and frequent reuse of their variants. We proposed to identify non-repeatable protein domains as the drug targets to disrupt the signal transduction than targeting the whole protein. Network based approach is used to understand the contribution of 52 domains in non-hub, non-essential, and intra-pathway cancerous nodes and to identify potential drug target domains. 34 distinct domains in the cancerous proteins are playing vital roles in making cancer as a complex disease and pose challenges to identify potential drug targets. Distribution of domain families follows the power law in the network. Single promiscuous domains are contributing to the formation of hubs like Pkinease, Pkinease Tyr, and Ras. Hub nodes are positively correlated with the domain coverage and targeting them would disrupt functional properties of the proteins. EIF 4EBP, alpha Kinase, Sel1, ROKNT, and KH 1 are the domains identified as potential domain targets for the disruption of the signaling mechanism involved in cancer.

Identification of genes with universally upregulated or downregulated expressions in colorectal cancer

BACKGROUND AND AIM

Differentially expressed (DE) genes detected at the population-level through case-control comparison provide no information on the dysregulation frequencies of DE genes in a cancer. In this work, we aimed to identify the genes with universally upregulated or downregulated expressions in colorectal cancer (CRC).

METHODS

We firstly clarified that DE genes in an individual cancer tissue should be the disease-relevant genes, which are dysregulated in the cancer tissue in comparison with its own previously normal state. Then, we identified DE genes at the individual level for 2233 CRC samples collected from multiple data sources using the RankComp algorithm.

RESULTS

We found 26 genes that were upregulated or downregulated in almost all the 2233 CRC samples and validated the results using 124 CRC tissues with paired adjacent normal tissues. Especially, we found that two genes (AJUBA and EGFL6), upregulated universally in CRC tissues, were extremely lowly expressed in normal colorectal tissues, which were considered to be oncogenes in CRC oncogenesis and development. Oppositely, we found that one gene (LPAR1), downregulated universally in CRC tissues, was silenced in CRC tissues but highly expressed in normal colorectal tissues, which were considered to be tumor suppressor genes in CRC. Functional evidences revealed that these three genes may induce CRC through deregulating pathways for ribosome biogenesis, cell proliferation, and cell cycle.

CONCLUSIONS

In conclusion, the individual-level DE genes analysis can help us find genes dysregulated universally in CRC tissues, which may be important diagnostic biomarkers and therapy targets.

A Pathway-Based Strategy to Identify Biomarkers for Lung Cancer Diagnosis and Prognosis

Current research has identified several potential biomarkers for lung cancer diagnosis or prognosis. However, most of these biomarkers are derived from a relatively small number of samples using algorithms at the gene level. Hence, gene expression signatures discovered in these studies have little overlaps. In this study, we proposed a new strategy to identify biomarkers from multiple datasets at the pathway level. We integrated the genome-wide expression data of lung cancer tissues from 13 published studies and applied our strategy to identify lung cancer diagnostic and prognostic biomarkers. We identified a 32-gene signature that differentiates lung adenocarcinomas from other lung cancer subtypes. We also discovered a 43-gene signature that can predict the outcome of human lung cancers. We tested their performance in several independent cohorts, which confirmed their robust prognostic and diagnostic power. Furthermore, we showed that the proposed gene expression signatures were independent of several traditional clinical indicators in lung cancer management. Our results suggest that the pathway-based strategy is useful to identify transcriptomic biomarkers from large-scale gene expression datasets that were collected from multiple sources.

Synthetic Lethality-based Identification of Targets for Anticancer Drugs in the Human Signaling Network

Chemotherapy agents can cause serious adverse effects by attacking both cancer tissues and normal tissues. Therefore, we proposed a synthetic lethality (SL) concept-based computational method to identify specific anticancer drug targets. First, a 3-step screening strategy (network-based, frequency-based and function-based screening) was proposed to identify the SL gene pairs by mining 697 cancer genes and the human signaling network, which had 6306 proteins and 62937 protein-protein interactions. The network-based screening was composed of a stability score constructed using a network information centrality measure (the average shortest path length) and the distance-based screening between the cancer gene and the non-cancer gene. Then, the non-cancer genes were extracted and annotated using drug-target interaction and drug description information to obtain potential anticancer drug targets. Finally, the human SL data in SynLethDB, the existing drug sensitivity data and text-mining were utilized for target validation. We successfully identified 2555 SL gene pairs and 57 potential anticancer drug targets. Among them, CDK1, CDK2, PLK1 and WEE1 were verified by all three aspects and could be preferentially used in specific targeted therapy in the future.

Identifying CpG sites with different differential methylation frequencies in colorectal cancer tissues based on individualized differential methylation analysis

A big challenge to clinical diagnosis and therapy of colorectal cancer (CRC) is its extreme heterogeneity, and thus it would be of special importance if we could find common biomarkers besides subtype-specific biomarkers for CRC. Here, with DNA methylation data produced by different laboratories, we firstly revealed that the relative methylation-level orderings (RMOs) of CpG sites within colorectal normal tissues are highly stable but widely disrupted in the CRC tissues. This finding provides the basis for using the RankComp algorithm to identify differentially methylated (DM) CpG sites in every individual CRC sample through comparing the RMOs within the individual sample with the stable RMOs predetermined in normal tissues. For 75 CRC samples, RankComp detected averagely 4,062 DM CpG sites per sample and reached an average precision of 91.34% in terms that the hypermethylation or hypomethylation states of the DM CpG sites detected for each cancer sample were consistent with the observed differences between this cancer sample and its paired adjacent normal sample. Finally, we applied RankComp to identify DM CpG sites for each of the 268 CRC samples from The Cancer Genome Atlas and found 26 and 143 genes whose promoter regions included CpG sites that were hypermethylated and hypomethylated, respectively, in more than 95% of the 268 CRC samples. Individualized pathway analysis identified six pathways that were significantly enriched with DM genes in more than 90% of the CRC tissues. These universal DNA methylation biomarkers could be important diagnostic makers and therapy targets for CRC.

Gene expression profiling analysis contributes to understanding the association between non-syndromic cleft lip and palate, and cancer

The present study aimed to investigate the molecular mechanisms underlying non-syndromic cleft lip, with or without cleft palate (NSCL/P), and the association between this disease and cancer. The GSE42589 data set was downloaded from the Gene Expression Omnibus database, and contained seven dental pulp stem cell samples from children with NSCL/P in the exfoliation period, and six controls. Differentially expressed genes (DEGs) were screened using the RankProd method, and their potential functions were revealed by pathway enrichment analysis and construction of a pathway interaction network. Subsequently, cancer genes were obtained from six cancer databases, and the cancer-associated protein-protein interaction network for the DEGs was visualized using Cytoscape. In total, 452 upregulated and 1,288 downregulated DEGs were screened. The upregulated DEGs were significantly enriched in the arachidonic acid metabolism pathway, including PTGDS, CYP4F2 and PLA2G16; and transforming growth factor (TGF)-β signaling pathway, including SMAD3 and TGFB2. The downregulated DEGs were distinctly involved in the pathways of DNA replication, including MCM2 and POLA1; cell cycle, including CDK1 and STAG1; and viral carcinogenesis, including PIK3CA and HIST1H2BF. Furthermore, the pathways of cell cycle and viral carcinogenesis, with higher degrees of interaction were found to interact with other pathways, including DNA replication, transcriptional misregulation in cancer, and the TGF-β signaling pathway. Additionally, TP53, CDK1, SMAD3, PIK3R1 and CASP3, with higher degrees, interacted with the cancer genes. In conclusion, the DEGs for NSCL/P were implicated predominantly in the TGF-β signaling pathway, the cell cycle and in viral carcinogenesis. The TP53, CDK1, SMAD3, PIK3R1 and CASP3 genes were found to be associated, not only with NSCL/P, but also with cancer. These results may contribute to a better understanding of the molecular mechanisms of NSCL/P.

Exploring the Functional Disorder and Corresponding Key Transcription Factors in Intraductal Papillary Mucinous Neoplasms Progression

This study has analyzed the gene expression patterns of an IPMN microarray dataset including normal pancreatic ductal tissue (NT), intraductal papillary mucinous adenoma (IPMA), intraductal papillary mucinous carcinoma (IPMC), and invasive ductal carcinoma (IDC) samples. And eight clusters of differentially expressed genes (DEGs) with similar expression pattern were detected by k-means clustering. Then a survey map of functional disorder in IPMN progression was established by functional enrichment analysis of these clusters. In addition, transcription factors (TFs) enrichment analysis was used to detect the key TFs in each cluster of DEGs, and three TFs (FLI1, ERG, and ESR1) were found to significantly regulate DEGs in cluster 1, and expression of these three TFs was validated by qRT-PCR. All these results indicated that these three TFs might play key roles in the early stages of IPMN progression.

Identifying network biomarkers based on protein-protein interactions and expression data

Identifying effective biomarkers to battle complex diseases is an important but challenging task in biomedical research today. Molecular data of complex diseases is increasingly abundant due to the rapid advance of high throughput technologies. However, a great gap remains in identifying the massive molecular data to phenotypic changes, in particular, at a network level, i.e., a novel method for identifying network biomarkers is in pressing need to accurately classify and diagnose diseases from molecular data and shed light on the mechanisms of disease pathogenesis. Rather than seeking differential genes at an individual-molecule level, here we propose a novel method for identifying network biomarkers based on protein-protein interaction affinity (PPIA), which identify the differential interactions at a network level. Specifically, we firstly define PPIAs by estimating the concentrations of protein complexes based on the law of mass action upon gene expression data. Then we select a small and non-redundant group of protein-protein interactions and single proteins according to the PPIAs, that maximizes the discerning ability of cases from controls. This method is mathematically formulated as a linear programming, which can be efficiently solved and guarantees a globally optimal solution. Extensive results on experimental data in breast cancer demonstrate the effectiveness and efficiency of the proposed method for identifying network biomarkers, which not only can accurately distinguish the phenotypes but also provides significant biological insights at a network or pathway level. In addition, our method provides a new way to integrate static protein-protein interaction information with dynamical gene expression data.

Integrated analysis of mutation data from various sources identifies key genes and signaling pathways in hepatocellular carcinoma

BACKGROUND

Recently, a number of studies have performed genome or exome sequencing of hepatocellular carcinoma (HCC) and identified hundreds or even thousands of mutations in protein-coding genes. However, these studies have only focused on a limited number of candidate genes, and many important mutation resources remain to be explored.

PRINCIPAL FINDINGS

In this study, we integrated mutation data obtained from various sources and performed pathway and network analysis. We identified 113 pathways that were significantly mutated in HCC samples and found that the mutated genes included in these pathways contained high percentages of known cancer genes, and damaging genes and also demonstrated high conservation scores, indicating their important roles in liver tumorigenesis. Five classes of pathways that were mutated most frequently included (a) proliferation and apoptosis related pathways, (b) tumor microenvironment related pathways, (c) neural signaling related pathways, (d) metabolic related pathways, and (e) circadian related pathways. Network analysis further revealed that the mutated genes with the highest betweenness coefficients, such as the well-known cancer genes TP53, CTNNB1 and recently identified novel mutated genes GNAL and the ADCY family, may play key roles in these significantly mutated pathways. Finally, we highlight several key genes (e.g., RPS6KA3 and PCLO) and pathways (e.g., axon guidance) in which the mutations were associated with clinical features.

CONCLUSIONS

Our workflow illustrates the increased statistical power of integrating multiple studies of the same subject, which can provide biological insights that would otherwise be masked under individual sample sets. This type of bioinformatics approach is consistent with the necessity of making the best use of the ever increasing data provided in valuable databases, such as TCGA, to enhance the speed of deciphering human cancers.

Pathway landscapes and epigenetic regulation in breast cancer and melanoma cell lines

BACKGROUND

Epigenetic variation is a main regulation mechanism of gene expression in various cancer histotypes, and due to its reversibility, the potential impact in therapy can be very relevant.

METHODS

Based on a selected pair, breast cancer (BC) and melanoma, we conducted inference analysis in parallel on a few cell lines (MCF-7 for BC and A375 for melanoma). Starting from differential expression after treatment with a demethylating agent, the 5-Aza-2'-deoxycytidine (DAC), we provided pathway enrichment analysis and gene regulatory maps with cross-linked microRNAs and transcription factors.

RESULTS

Several oncogenic signaling pathways altered upon DAC treatment were detected with significant enrichment. We represented the association between these cancers by depicting the landscape of common and specific variation affecting them.

Genes dysregulated to different extent or oppositely in estrogen receptor-positive and estrogen receptor-negative breast cancers

Background

Directly comparing gene expression profiles of estrogen receptor-positive (ER+) and estrogen receptor-negative (ER−) breast cancers cannot determine whether differentially expressed genes between these two subtypes result from dysregulated expression in ER+ cancer or ER− cancer versus normal controls, and thus would miss critical information for elucidating the transcriptomic difference between the two subtypes.

Principal Findings

Using microarray datasets from TCGA, we classified the genes dysregulated in both ER+ and ER− cancers versus normal controls into two classes: (i) genes dysregulated in the same direction but to a different extent, and (ii) genes dysregulated to opposite directions, and then validated the two classes in RNA-sequencing datasets of independent cohorts. We showed that the genes dysregulated to a larger extent in ER+ cancers than in ER− cancers enriched in glycerophospholipid and polysaccharide metabolic processes, while the genes dysregulated to a larger extent in ER− cancers than in ER+ cancers enriched in cell proliferation. Phosphorylase kinase and enzymes of glycosylphosphatidylinositol (GPI) anchor biosynthesis were upregulated to a larger extent in ER+ cancers than in ER− cancers, whereas glycogen synthase and phospholipase A2 were downregulated to a larger extent in ER+ cancers than in ER− cancers. We also found that the genes oppositely dysregulated in the two subtypes significantly enriched with known cancer genes and tended to closely collaborate with the cancer genes. Furthermore, we showed the possibility that these oppositely dysregulated genes could contribute to carcinogenesis of ER+ and ER− cancers through rewiring different subpathways.

Conclusions

GPI-anchor biosynthesis and glycogenolysis were elevated and hydrolysis of phospholipids was depleted to a larger extent in ER+ cancers than in ER− cancers. Our findings indicate that the genes oppositely dysregulated in the two subtypes are potential cancer genes which could contribute to carcinogenesis of both ER+ and ER− cancers through rewiring different subpathways.

Warehousing re-annotated cancer genes for biomarker meta-analysis

Translational research in cancer genomics assigns a fundamental role to bioinformatics in support of candidate gene prioritization with regard to both biomarker discovery and target identification for drug development. Efforts in both such directions rely on the existence and constant update of large repositories of gene expression data and omics records obtained from a variety of experiments. Users who interactively interrogate such repositories may have problems in retrieving sample fields that present limited associated information, due for instance to incomplete entries or sometimes unusable files. Cancer-specific data sources present similar problems. Given that source integration usually improves data quality, one of the objectives is keeping the computational complexity sufficiently low to allow an optimal assimilation and mining of all the information. In particular, the scope of integrating intraomics data can be to improve the exploration of gene co-expression landscapes, while the scope of integrating interomics sources can be that of establishing genotype-phenotype associations. Both integrations are relevant to cancer biomarker meta-analysis, as the proposed study demonstrates. Our approach is based on re-annotating cancer-specific data available at the EBI's ArrayExpress repository and building a data warehouse aimed to biomarker discovery and validation studies. Cancer genes are organized by tissue with biomedical and clinical evidences combined to increase reproducibility and consistency of results. For better comparative evaluation, multiple queries have been designed to efficiently address all types of experiments and platforms, and allow for retrieval of sample-related information, such as cell line, disease state and clinical aspects.

Extracting a few functionally reproducible biomarkers to build robust subnetwork-based classifiers for the diagnosis of cancer

In microarray-based case-control studies of a disease, people often attempt to identify a few diagnostic or prognostic markers amongst the most significant differentially expressed (DE) genes. However, the reproducibility of DE genes identified in different studies for a disease is typically very low. To tackle the problem, we could evaluate the reproducibility of DE genes across studies and define robust markers for disease diagnosis using disease-associated protein-protein interaction (PPI) subnetwork. Using datasets for four cancer types, we found that the most significant DE genes in cancer exhibit consistent up- or down-regulation in different datasets. For each cancer type, the 5 (or 10) most significant DE genes separately extracted from different datasets tend to be significantly coexpressed and closely connected in the PPI subnetwork, thereby indicating that they are highly reproducible at the PPI level. Consequently, we were able to build robust subnetwork-based classifiers for cancer diagnosis.

An integrated approach to uncover driver genes in breast cancer methylation genomes

Background

Cancer cells typically exhibit large-scale aberrant methylation of gene promoters. Some of the genes with promoter methylation alterations play “driver” roles in tumorigenesis, whereas others are only “passengers”.

Results

Based on the assumption that promoter methylation alteration of a driver gene may lead to expression alternation of a set of genes associated with cancer pathways, we developed a computational framework for integrating promoter methylation and gene expression data to identify driver methylation aberrations of cancer. Applying this approach to breast cancer data, we identified many novel cancer driver genes and found that some of the identified driver genes were subtype-specific for basal-like, luminal-A and HER2+ subtypes of breast cancer.

Conclusion

The proposed framework proved effective in identifying cancer driver genes from genome-wide gene methylation and expression data of cancer. These results may provide new molecular targets for potential targeted and selective epigenetic therapy.

Pitfalls in experimental designs for characterizing the transcriptional, methylational and copy number changes of oncogenes and tumor suppressor genes

BACKGROUND

It is a common practice that researchers collect a set of samples without discriminating the mutants and their wild-type counterparts to characterize the transcriptional, methylational and/or copy number changes of pre-defined candidate oncogenes or tumor suppressor genes (TSGs), although some examples are known that carcinogenic mutants may express and function completely differently from their wild-type counterparts.

PRINCIPAL FINDINGS

Based on various high-throughput data without mutation information for typical cancer types, we surprisingly found that about half of known oncogenes (or TSGs) pre-defined by mutations were down-regulated (or up-regulated) and hypermethylated (or hypomethylated) in their corresponding cancer types. Therefore, the overall expression and/or methylation changes of genes detected in a set of samples without discriminating the mutants and their wild-type counterparts cannot indicate the carcinogenic roles of the mutants. We also found that about half of known oncogenes were located in deletion regions, whereas all known TSGs were located in deletion regions. Thus, both oncogenes and TSGs may be located in deletion regions and thus deletions can indicate TSGs only if the gene is found to be deleted as a whole. In contrast, amplifications are restricted to oncogenes and thus can be used to support either the dysregulated wild-type gene or its mutant as an oncogene.

CONCLUSIONS

We demonstrated that using the transcriptional, methylational and/or copy number changes without mutation information to characterize oncogenes and TSGs, which is a currently still widely adopted strategy, will most often produce misleading results. Our analysis highlights the importance of evaluating expression, methylation and copy number changes together with gene mutation data in the same set of samples in order to determine the distinct roles of the mutants and their wild-type counterparts.

Network analysis of genomic alteration profiles reveals co-altered functional modules and driver genes for glioblastoma

The heterogeneity of genetic alterations in human cancer genomes presents a major challenge to advancing our understanding of cancer mechanisms and identifying cancer driver genes. To tackle this heterogeneity problem, many approaches have been proposed to investigate genetic alterations and predict driver genes at the individual pathway level. However, most of these approaches ignore the correlation of alteration events between pathways and miss many genes with rare alterations collectively contributing to carcinogenesis. Here, we devise a network-based approach to capture the cooperative functional modules hidden in genome-wide somatic mutation and copy number alteration profiles of glioblastoma (GBM) from The Cancer Genome Atlas (TCGA), where a module is a set of altered genes with dense interactions in the protein interaction network. We identify 7 pairs of significantly co-altered modules that involve the main pathways known to be altered in GBM (TP53, RB and RTK signaling pathways) and highlight the striking co-occurring alterations among these GBM pathways. By taking into account the non-random correlation of gene alterations, the property of co-alteration could distinguish oncogenic modules that contain driver genes involved in the progression of GBM. The collaboration among cancer pathways suggests that the redundant models and aggravating models could shed new light on the potential mechanisms during carcinogenesis and provide new indications for the design of cancer therapeutic strategies.

Distinct and competitive regulatory patterns of tumor suppressor genes and oncogenes in ovarian cancer

Background

So far, investigators have found numerous tumor suppressor genes (TSGs) and oncogenes (OCGs) that control cell proliferation and apoptosis during cancer development. Furthermore, TSGs and OCGs may act as modulators of transcription factors (TFs) to influence gene regulation. A comprehensive investigation of TSGs, OCGs, TFs, and their joint target genes at the network level may provide a deeper understanding of the post-translational modulation of TSGs and OCGs to TF gene regulation.

Methodology/Principal Findings

In this study, we developed a novel computational framework for identifying target genes of TSGs and OCGs using TFs as bridges through the integration of protein-protein interactions and gene expression data. We applied this pipeline to ovarian cancer and constructed a three-layer regulatory network. In the network, the top layer was comprised of modulators (TSGs and OCGs), the middle layer included TFs, and the bottom layer contained target genes. Based on regulatory relationships in the network, we compiled TSG and OCG profiles and performed clustering analyses. Interestingly, we found TSGs and OCGs formed two distinct branches. The genes in the TSG branch were significantly enriched in DNA damage and repair, regulating macromolecule metabolism, cell cycle and apoptosis, while the genes in the OCG branch were significantly enriched in the ErbB signaling pathway. Remarkably, their specific targets showed a reversed functional enrichment in terms of apoptosis and the ErbB signaling pathway: the target genes regulated by OCGs only were enriched in anti-apoptosis and the target genes regulated by TSGs only were enriched in the ErbB signaling pathway.

Conclusions/Significance

This study provides the first comprehensive investigation of the interplay of TSGs and OCGs in a regulatory network modulated by TFs. Our application in ovarian cancer revealed distinct regulatory patterns of TSGs and OCGs, suggesting a competitive regulatory mechanism acting upon apoptosis and the ErbB signaling pathway through their specific target genes.

Single sample expression-anchored mechanisms predict survival in head and neck cancer

Gene expression signatures that are predictive of therapeutic response or prognosis are increasingly useful in clinical care; however, mechanistic (and intuitive) interpretation of expression arrays remains an unmet challenge. Additionally, there is surprisingly little gene overlap among distinct clinically validated expression signatures. These “causality challenges” hinder the adoption of signatures as compared to functionally well-characterized single gene biomarkers. To increase the utility of multi-gene signatures in survival studies, we developed a novel approach to generate “personal mechanism signatures” of molecular pathways and functions from gene expression arrays. FAIME, the Functional Analysis of Individual Microarray Expression, computes mechanism scores using rank-weighted gene expression of an individual sample. By comparing head and neck squamous cell carcinoma (HNSCC) samples with non-tumor control tissues, the precision and recall of deregulated FAIME-derived mechanisms of pathways and molecular functions are comparable to those produced by conventional cohort-wide methods (e.g. GSEA). The overlap of “Oncogenic FAIME Features of HNSCC” (statistically significant and differentially regulated FAIME-derived genesets representing GO functions or KEGG pathways derived from HNSCC tissue) among three distinct HNSCC datasets (pathways:46%, p<0.001) is more significant than the gene overlap (genes:4%). These Oncogenic FAIME Features of HNSCC can accurately discriminate tumors from control tissues in two additional HNSCC datasets (n = 35 and 91, F-accuracy = 100% and 97%, empirical p<0.001, area under the receiver operating characteristic curves = 99% and 92%), and stratify recurrence-free survival in patients from two independent studies (p = 0.0018 and p = 0.032, log-rank). Previous approaches depending on group assignment of individual samples before selecting features or learning a classifier are limited by design to discrete-class prediction. In contrast, FAIME calculates mechanism profiles for individual patients without requiring group assignment in validation sets. FAIME is more amenable for clinical deployment since it translates the gene-level measurements of each given sample into pathways and molecular function profiles that can be applied to analyze continuous phenotypes in clinical outcome studies (e.g. survival time, tumor volume).

Clinical utilization of multi-gene expression signatures that are predictive of therapeutic response has been steadily increasing, however, interpretation of such results remains challenging because multi-gene signatures, generated from analyzing different patient cohorts, tend to be equally predictive but contain minimal overlap. Whereas pathway-level analyses of expression arrays show promise for generating clinically meaningful mechanistic signatures, current approaches do not permit single-patient based analyses that are independent of cross-group calculations. To bridge the gap between deterministic biological mechanisms of single-gene biomarkers and the statistical predictive power of multi-gene signatures that are disconnected from mechanisms, we developed FAIME, a novel method that transforms microarray gene expression data into individualized patient profiles of molecular mechanisms. We have validated its capability for predicting clinical outcomes, including cancer patient samples derived from six different clinical trial cohorts of head and neck cancers. This method provides opportunities to harness an untapped resource for personal genomics: clinical evaluation and testing of individually interpretable mechanistic profiles derived from gene expression arrays.

Reproducibility and concordance of differential DNA methylation and gene expression in cancer

Background

Hundreds of genes with differential DNA methylation of promoters have been identified for various cancers. However, the reproducibility of differential DNA methylation discoveries for cancer and the relationship between DNA methylation and aberrant gene expression have not been systematically analysed.

Methodology/Principal Findings

Using array data for seven types of cancers, we first evaluated the effects of experimental batches on differential DNA methylation detection. Second, we compared the directions of DNA methylation changes detected from different datasets for the same cancer. Third, we evaluated the concordance between methylation and gene expression changes. Finally, we compared DNA methylation changes in different cancers. For a given cancer, the directions of methylation and expression changes detected from different datasets, excluding potential batch effects, were highly consistent. In different cancers, DNA hypermethylation was highly inversely correlated with the down-regulation of gene expression, whereas hypomethylation was only weakly correlated with the up-regulation of genes. Finally, we found that genes commonly hypomethylated in different cancers primarily performed functions associated with chronic inflammation, such as ‘keratinization’, ‘chemotaxis’ and ‘immune response’.

Conclusions

Batch effects could greatly affect the discovery of DNA methylation biomarkers. For a particular cancer, both differential DNA methylation and gene expression can be reproducibly detected from different studies with no batch effects. While DNA hypermethylation is significantly linked to gene down-regulation, hypomethylation is only weakly correlated with gene up-regulation and is likely to be linked to chronic inflammation.

Network based consensus gene signatures for biomarker discovery in breast cancer

Diagnostic and prognostic biomarkers for cancer based on gene expression profiles are viewed as a major step towards a better personalized medicine. Many studies using various computational approaches have been published in this direction during the last decade. However, when comparing different gene signatures for related clinical questions often only a small overlap is observed. This can have various reasons, such as technical differences of platforms, differences in biological samples or their treatment in lab, or statistical reasons because of the high dimensionality of the data combined with small sample size, leading to unstable selection of genes. In conclusion retrieved gene signatures are often hard to interpret from a biological point of view. We here demonstrate that it is possible to construct a consensus signature from a set of seemingly different gene signatures by mapping them on a protein interaction network. Common upstream proteins of close gene products, which we identified via our developed algorithm, show a very clear and significant functional interpretation in terms of overrepresented KEGG pathways, disease associated genes and known drug targets. Moreover, we show that such a consensus signature can serve as prior knowledge for predictive biomarker discovery in breast cancer. Evaluation on different datasets shows that signatures derived from the consensus signature reveal a much higher stability than signatures learned from all probesets on a microarray, while at the same time being at least as predictive. Furthermore, they are clearly interpretable in terms of enriched pathways, disease associated genes and known drug targets. In summary we thus believe that network based consensus signatures are not only a way to relate seemingly different gene signatures to each other in a functional manner, but also to establish prior knowledge for highly stable and interpretable predictive biomarkers.

Reproducible cancer biomarker discovery in SELDI-TOF MS using different pre-processing algorithms

Background

There has been much interest in differentiating diseased and normal samples using biomarkers derived from mass spectrometry (MS) studies. However, biomarker identification for specific diseases has been hindered by irreproducibility. Specifically, a peak profile extracted from a dataset for biomarker identification depends on a data pre-processing algorithm. Until now, no widely accepted agreement has been reached.

Results

In this paper, we investigated the consistency of biomarker identification using differentially expressed (DE) peaks from peak profiles produced by three widely used average spectrum-dependent pre-processing algorithms based on SELDI-TOF MS data for prostate and breast cancers. Our results revealed two important factors that affect the consistency of DE peak identification using different algorithms. One factor is that some DE peaks selected from one peak profile were not detected as peaks in other profiles, and the second factor is that the statistical power of identifying DE peaks in large peak profiles with many peaks may be low due to the large scale of the tests and small number of samples. Furthermore, we demonstrated that the DE peak detection power in large profiles could be improved by the stratified false discovery rate (FDR) control approach and that the reproducibility of DE peak detection could thereby be increased.

Conclusions

Comparing and evaluating pre-processing algorithms in terms of reproducibility can elucidate the relationship among different algorithms and also help in selecting a pre-processing algorithm. The DE peaks selected from small peak profiles with few peaks for a dataset tend to be reproducibly detected in large peak profiles, which suggests that a suitable pre-processing algorithm should be able to produce peaks sufficient for identifying useful and reproducible biomarkers.

Learning biomarkers of pluripotent stem cells in mouse

Pluripotent stem cells are able to self-renew, and to differentiate into all adult cell types. Many studies report data describing these cells, and characterize them in molecular terms. Machine learning yields classifiers that can accurately identify pluripotent stem cells, but there is a lack of studies yielding minimal sets of best biomarkers (genes/features). We assembled gene expression data of pluripotent stem cells and non-pluripotent cells from the mouse. After normalization and filtering, we applied machine learning, classifying samples into pluripotent and non-pluripotent with high cross-validated accuracy. Furthermore, to identify minimal sets of best biomarkers, we used three methods: information gain, random forests and a wrapper of genetic algorithm and support vector machine (GA/SVM). We demonstrate that the GA/SVM biomarkers work best in combination with each other; pathway and enrichment analyses show that they cover the widest variety of processes implicated in pluripotency. The GA/SVM wrapper yields best biomarkers, no matter which classification method is used. The consensus best biomarker based on the three methods is Tet1, implicated in pluripotency just recently. The best biomarker based on the GA/SVM wrapper approach alone is Fam134b, possibly a missing link between pluripotency and some standard surface markers of unknown function processed by the Golgi apparatus.

Human protein reference database and human proteinpedia as discovery resources for molecular biotechnology

In the recent years, research in molecular biotechnology has transformed from being small scale studies targeted at a single or a small set of molecule(s) into a combination of high throughput discovery platforms and extensive validations. Such a discovery platform provided an unbiased approach which resulted in the identification of several novel genetic and protein biomarkers. High throughput nature of these investigations coupled with higher sensitivity and specificity of Next Generation technologies provided qualitatively and quantitatively richer biological data. These developments have also revolutionized biological research and speed of data generation. However, it is becoming difficult for individual investigators to directly benefit from this data because they are not easily accessible. Data resources became necessary to assimilate, store and disseminate information that could allow future discoveries. We have developed two resources--Human Protein Reference Database (HPRD) and Human Proteinpedia, which integrate knowledge relevant to human proteins. A number of protein features including protein-protein interactions, post-translational modifications, subcellular localization, and tissue expression, which have been studied using different strategies were incorporated in these databases. Human Proteinpedia also provides a portal for community participation to annotate and share proteomic data and uses HPRD as the scaffold for data processing. Proteomic investigators can even share unpublished data in Human Proteinpedia, which provides a meaningful platform for data sharing. As proteomic information reflects a direct view of cellular systems, proteomics is expected to complement other areas of biology such as genomics, transcriptomics, molecular biology, cloning, and classical genetics in understanding the relationships among multiple facets of biological systems.

Analysis of pathway mutation profiles highlights collaboration between cancer-associated superpathways

The biological interpretation of the complexity of cancer somatic mutation profiles is a major challenge in current cancer research. It has been suggested that mutations in multiple genes that participate in different pathways are collaborative in conferring growth advantage to tumor cells. Here, we propose a powerful pathway-based approach to study the functional collaboration of gene mutations in carcinogenesis. We successfully identify many pairs of significantly comutated pathways for a large-scale somatic mutation profile of lung adenocarcinoma. We find that the coordinated pathway pairs detected by comutations are also likely to be coaltered by other molecular changes, such as alterations in multifunctional genes in cancer. Then, we cluster comutated pathways into comutated superpathways and show that the derived superpathways also tend to be significantly coaltered by DNA copy number alterations. Our results support the hypothesis that comprehensive cooperation among a few basic functions is required for inducing cancer. The results also suggest biologically plausible models for understanding the heterogeneous mechanisms of cancers. Finally, we suggest an approach to identify candidate cancer genes from the derived comutated pathways. Together, our results provide guidelines to distill the pathway collaboration in carcinogenesis from the complexity of cancer somatic mutation profiles.

Evaluating the consistency of differential expression of microRNA detected in human cancers

Differential expression of microRNA (miRNA) is involved in many human diseases and could potentially be used as a biomarker for disease diagnosis, prognosis, and therapy. However, inconsistency has often been found among differentially expressed miRNAs identified in various studies when using miRNA arrays for a particular disease such as a cancer. Before broadly applying miRNA arrays in a clinical setting, it is critical to evaluate inconsistent discoveries in a rational way. Thus, using data sets from 2 types of cancers, our study shows that the differentially expressed miRNAs detected from multiple experiments for each cancer exhibit stable regulation direction. This result also indicates that miRNA arrays could be used to reliably capture the signals of the regulation direction of differentially expressed miRNAs in cancer. We then assumed that 2 differentially expressed miRNAs with the same regulation direction in a particular cancer play similar functional roles if they regulate the same set of cancer-associated genes. On the basis of this hypothesis, we proposed a score to assess the functional consistency between differentially expressed miRNAs separately extracted from multiple studies for a particular cancer. We showed although lists of differentially expressed miRNAs identified from different studies for each cancer were highly variable, they were rather consistent at the level of function. Thus, the detection of differentially expressed miRNAs in various experiments for a certain disease tends to be functionally reproducible and capture functionally related differential expression of miRNAs in the disease.

Multi-level reproducibility of signature hubs in human interactome for breast cancer metastasis

BACKGROUND

It has been suggested that, in the human protein-protein interaction network, changes of co-expression between highly connected proteins ("hub") and their interaction neighbours might have important roles in cancer metastasis and be predictive disease signatures for patient outcome. However, for a cancer, such disease signatures identified from different studies have little overlap.

RESULTS

Here, we propose a systemic approach to evaluate the reproducibility of disease signatures at multiple levels, on the basis of some statistically testable biological models. Using two datasets for breast cancer metastasis, we showed that different signature hubs identified from different studies were highly consistent in terms of significantly sharing interaction neighbours and displaying consistent co-expression changes with their overlapping neighbours, whereas the shared interaction neighbours were significantly over-represented with known cancer genes and enriched in pathways deregulated in breast cancer pathogenesis. Then, we showed that the signature hubs identified from the two datasets were highly reproducible at the protein interaction and pathway levels in three other independent datasets.

CONCLUSIONS

Our results provide a possible biological model that different signature hubs altered in different patient cohorts could disturb the same pathways associated with cancer metastasis through their interaction neighbours.

Systematic interpretation of comutated genes in large-scale cancer mutation profiles

By high-throughput screens of somatic mutations of genes in cancer genomes, hundreds of cancer genes are being rapidly identified, providing us abundant information for systematically deciphering the genetic changes underlying cancer mechanism. However, the functional collaboration of mutated genes is often neglected in current studies. Here, using four genome-wide somatic mutation data sets and pathways defined in various databases, we showed that gene pairs significantly comutated in cancer samples tend to distribute between pathways rather than within pathways. At the basic functional level of motifs in the human protein-protein interaction network, we also found that comutated gene pairs were overrepresented between motifs but extremely depleted within motifs. Specifically, we showed that based on Gene Ontology that describes gene functions at various specific levels, we could tackle the pathway definition problem to some degree and study the functional collaboration of gene mutations in cancer genomes more efficiently. Then, by defining pairs of pathways frequently linked by comutated gene pairs as the between-pathway models, we showed they are also likely to be codisrupted by mutations of the interpathway hubs of the coupled pathways, suggesting new hints for understanding the heterogeneous mechanisms of cancers. Finally, we showed some between-pathway models consisting of important pathways such as cell cycle checkpoint and cell proliferation were codisrupted in most cancer samples under this study, suggesting that their codisruptions might be functionally essential in inducing these cancers. All together, our results would provide a channel to detangle the complex collaboration of the molecular processes underlying cancer mechanism.