Tumor characterization and stratification by integrated molecular profiles reveals essential pan-cancer features

Leveraging technology-driven strategies to untangle omics big data: circumventing roadblocks in clinical facets of oral cancer

Oral cancer is one of the 19most rapidly progressing cancers associated with significant mortality, owing to its extreme degree of invasiveness and aggressive inclination. The early occurrences of this cancer can be clinically deceiving leading to a poor overall survival rate. The primary concerns from a clinical perspective include delayed diagnosis, rapid disease progression, resistance to various chemotherapeutic regimens, and aggressive metastasis, which collectively pose a substantial threat to prognosis. Conventional clinical practices observed since antiquity no longer offer the best possible options to circumvent these roadblocks. The world of current cancer research has been revolutionized with the advent of state-of-the-art technology-driven strategies that offer a ray of hope in confronting said challenges by highlighting the crucial underlying molecular mechanisms and drivers. In recent years, bioinformatics and Machine Learning (ML) techniques have enhanced the possibility of early detection, evaluation of prognosis, and individualization of therapy. This review elaborates on the application of the aforesaid techniques in unraveling potential hints from omics big data to address the complexities existing in various clinical facets of oral cancer. The first section demonstrates the utilization of omics data and ML to disentangle the impediments related to diagnosis. This includes the application of technology-based strategies to optimize early detection, classification, and staging via uncovering biomarkers and molecular signatures. Furthermore, breakthrough concepts such as salivaomics-driven non-invasive biomarker discovery and omics-complemented surgical interventions are articulated in detail. In the following part, the identification of novel disease-specific targets alongside potential therapeutic agents to confront oral cancer via omics-based methodologies is presented. Additionally, a special emphasis is placed on drug resistance, precision medicine, and drug repurposing. In the final section, we discuss the research approaches oriented toward unveiling the prognostic biomarkers and constructing prediction models to capture the metastatic potential of the tumors. Overall, we intend to provide a bird's eye view of the various omics, bioinformatics, and ML approaches currently being used in oral cancer research through relevant case studies.

Multi-Omics Data Analysis Identifies Prognostic Biomarkers across Cancers

Combining omics data from different layers using integrative methods provides a better understanding of the biology of a complex disease such as cancer. The discovery of biomarkers related to cancer development or prognosis helps to find more effective treatment options. This study integrates multi-omics data of different cancer types with a network-based approach to explore common gene modules among different tumors by running community detection methods on the integrated network. The common modules were evaluated by several biological metrics adapted to cancer. Then, a new prognostic scoring method was developed by weighting mRNA expression, methylation, and mutation status of genes. The survival analysis pointed out statistically significant results for GNG11, CBX2, CDKN3, ARHGEF10, CLN8, SEC61G and PTDSS1 genes. The literature search reveals that the identified biomarkers are associated with the same or different types of cancers. Our method does not only identify known cancer-specific biomarker genes, but also proposes new potential biomarkers. Thus, this study provides a rationale for identifying new gene targets and expanding treatment options across cancer types.

Network diffusion-based approach for survival prediction and identification of biomarkers using multi-omics data of papillary renal cell carcinoma

Identification of cancer subtypes based on molecular knowledge is crucial for improving the patient diagnosis, prognosis, and treatment. In this work, we integrated copy number variations (CNVs) and transcriptomic data of Kidney Papillary Renal Cell Carcinoma (KIRP) using a network diffusion strategy to stratify cancers into clinically and biologically relevant subtypes. We constructed GeneNet, a KIRP specific gene expression network from RNA-seq data. The copy number variation data was projected onto GeneNet and propagated on the network for clustering. We identified robust subtypes that are biologically informative and significantly associated with patient survival, tumor stage and clinical subtypes of KIRP. We performed a Singular Value Decomposition (SVD) analysis of KIRP subtypes, which revealed the genes/silent players related to poor survival. A differential gene expression analysis between subtypes showed that genes related to immune, extracellular matrix organization, and genomic instability are upregulated in the poor survival group. Overall, the network-based approach revealed the molecular subtypes of KIRP and captured the relationship between gene expression and CNVs. This framework can be further expanded to integrate other omics data.

An integrated analysis of the cancer genome atlas data discovers a hierarchical association structure across thirty three cancer types

Cancer cells harbor molecular alterations at all levels of information processing. Genomic/epigenomic and transcriptomic alterations are inter-related between genes, within and across cancer types and may affect clinical phenotypes. Despite the abundant prior studies of integrating cancer multi-omics data, none of them organizes these associations in a hierarchical structure and validates the discoveries in extensive external data. We infer this Integrated Hierarchical Association Structure (IHAS) from the complete data of The Cancer Genome Atlas (TCGA) and compile a compendium of cancer multi-omics associations. Intriguingly, diverse alterations on genomes/epigenomes from multiple cancer types impact transcriptions of 18 Gene Groups. Half of them are further reduced to three Meta Gene Groups enriched with (1) immune and inflammatory responses, (2) embryonic development and neurogenesis, (3) cell cycle process and DNA repair. Over 80% of the clinical/molecular phenotypes reported in TCGA are aligned with the combinatorial expressions of Meta Gene Groups, Gene Groups, and other IHAS subunits. Furthermore, IHAS derived from TCGA is validated in more than 300 external datasets including multi-omics measurements and cellular responses upon drug treatments and gene perturbations in tumors, cancer cell lines, and normal tissues. To sum up, IHAS stratifies patients in terms of molecular signatures of its subunits, selects targeted genes or drugs for precision cancer therapy, and demonstrates that associations between survival times and transcriptional biomarkers may vary with cancer types. These rich information is critical for diagnosis and treatments of cancers.

Discovery of pan-cancer related genes via integrative network analysis

Identification of cancer-related genes is helpful for understanding the pathogenesis of cancer, developing targeted drugs and creating new diagnostic and therapeutic methods. Considering the complexity of the biological laboratory methods, many network-based methods have been proposed to identify cancer-related genes at the global perspective with the increasing availability of high-throughput data. Some studies have focused on the tissue-specific cancer networks. However, cancers from different tissues may share common features, and those methods may ignore the differences and similarities across cancers during the establishment of modeling. In this work, in order to make full use of global information of the network, we first establish the pan-cancer network via differential network algorithm, which not only contains heterogeneous data across multiple cancer types but also contains heterogeneous data between tumor samples and normal samples. Second, the node representation vectors are learned by network embedding. In contrast to ranking analysis-based methods, with the help of integrative network analysis, we transform the cancer-related gene identification problem into a binary classification problem. The final results are obtained via ensemble classification. We further applied these methods to the most commonly used gene expression data involving six tissue-specific cancer types. As a result, an integrative pan-cancer network and several biologically meaningful results were obtained. As examples, nine genes were ultimately identified as potential pan-cancer-related genes. Most of these genes have been reported in published studies, thus showing our method's potential for application in identifying driver gene candidates for further biological experimental verification.

From multi-omics integration towards novel genomic interaction networks to identify key cancer cell line characteristics

Cancer is a complex disease where cancer cells express epigenetic and transcriptomic mechanisms to promote tumor initiation, progression, and survival. To extract relevant features from the 2019 Cancer Cell Line Encyclopedia (CCLE), a multi-layer nonnegative matrix factorization approach is used. We used relevant feature genes and DNA promoter regions to construct genomic interaction network to study gene-gene and gene-DNA promoter methylation relationships. Here, we identified a set of gene transcripts and methylated DNA promoter regions for different clusters, including one homogeneous lymphoid neoplasms cluster. In this cluster, we found different methylated transcription factors that affect transcriptional activation of EGFR and downstream interactions. Furthermore, the hippo-signaling pathway might not function properly because of DNA hypermethylation and low gene expression of both LATS2 and YAP1. Finally, we could identify a potential dysregulation of the CD28-CD86-CTLA4 axis. Characterizing the interaction of the epigenome and the transcriptome is vital for our understanding of cancer cell line behavior, not only for deepening insights into cancer-related processes but also for future disease treatment and drug development. Here we have identified potential candidates that characterize cancer cell lines, which give insight into the development and progression of cancers.

Identifying modules of cooperating cancer drivers

Identifying cooperating modules of driver alterations can provide insights into cancer etiology and advance the development of effective personalized treatments. We present Cancer Rule Set Optimization (CRSO) for inferring the combinations of alterations that cooperate to drive tumor formation in individual patients. Application to 19 TCGA cancer types revealed a mean of 11 core driver combinations per cancer, comprising 2-6 alterations per combination and accounting for a mean of 70% of samples per cancer type. CRSO is distinct from methods based on statistical co-occurrence, which we demonstrate is a suboptimal criterion for investigating driver cooperation. CRSO identified well-studied driver combinations that were not detected by other approaches and nominated novel combinations that correlate with clinical outcomes in multiple cancer types. Novel synergies were identified in NRAS-mutant melanomas that may be therapeutically relevant. Core driver combinations involving NFE2L2 mutations were identified in four cancer types, supporting the therapeutic potential of NRF2 pathway inhibition. CRSO is available at https://github.com/mikekleinsgit/CRSO/.

ETumorMetastasis: A Network-based Algorithm Predicts Clinical Outcomes Using Whole-exome Sequencing Data of Cancer Patients

Continual reduction in sequencing cost is expanding the accessibility of genome sequencing data for routine clinical applications. However, the lack of methods to construct machine learning-based predictive models using these datasets has become a crucial bottleneck for the application of sequencing technology in clinics. Here, we develop a new algorithm, eTumorMetastasis, which transforms tumor functional mutations into network-based profiles and identifies network operational gene (NOG) signatures. NOG signatures model the tipping point at which a tumor cell shifts from a state that doesn’t favor recurrence to one that does. We show that NOG signatures derived from genomic mutations of tumor founding clones (i.e., the ‘most recent common ancestor’ of the cells within a tumor) significantly distinguish the recurred and non-recurred breast tumors as well as outperform the most popular genomic test (i.e., Oncotype DX). These results imply that mutations of the tumor founding clones are associated with tumor recurrence and can be used to predict clinical outcomes. As such, predictive tools could be used in clinics to guide treatment routes. Finally, the concepts underlying the eTumorMetastasis pave the way for the application of genome sequencing in predictions for other complex genetic diseases. eTumorMetastasis pseudocode and related data used in this study are available at https://github.com/WangEdwinLab/eTumorMetastasis.

Inferring subgroup-specific driver genes from heterogeneous cancer samples via subspace learning with subgroup indication

MOTIVATION

Detecting driver genes from gene mutation data is a fundamental task for tumorigenesis research. Due to the fact that cancer is a heterogeneous disease with various subgroups, subgroup-specific driver genes are the key factors in the development of precision medicine for heterogeneous cancer. However, the existing driver gene detection methods are not designed to identify subgroup specificities of their detected driver genes, and therefore cannot indicate which group of patients is associated with the detected driver genes, which is difficult to provide specifically clinical guidance for individual patients.

RESULTS

By incorporating the subspace learning framework, we propose a novel bioinformatics method called DriverSub, which can efficiently predict subgroup-specific driver genes in the situation where the subgroup annotations are not available. When evaluated by simulation datasets with known ground truth and compared with existing methods, DriverSub yields the best prediction of driver genes and the inference of their related subgroups. When we apply DriverSub on the mutation data of real heterogeneous cancers, we can observe that the predicted results of DriverSub are highly enriched for experimentally validated known driver genes. Moreover, the subgroups inferred by DriverSub are significantly associated with the annotated molecular subgroups, indicating its capability of predicting subgroup-specific driver genes.

AVAILABILITY AND IMPLEMENTATION

The source code is publicly available at https://github.com/JianingXi/DriverSub.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Network Diffusion Promotes the Integrative Analysis of Multiple Omics

The development of integrative methods is one of the main challenges in bioinformatics. Network-based methods for the analysis of multiple gene-centered datasets take into account known and/or inferred relations between genes. In the last decades, the mathematical machinery of network diffusion—also referred to as network propagation—has been exploited in several network-based pipelines, thanks to its ability of amplifying association between genes that lie in network proximity. Indeed, network diffusion provides a quantitative estimation of network proximity between genes associated with one or more different data types, from simple binary vectors to real vectors. Therefore, this powerful data transformation method has also been increasingly used in integrative analyses of multiple collections of biological scores and/or one or more interaction networks. We present an overview of the state of the art of bioinformatics pipelines that use network diffusion processes for the integrative analysis of omics data. We discuss the fundamental ways in which network diffusion is exploited, open issues and potential developments in the field. Current trends suggest that network diffusion is a tool of broad utility in omics data analysis. It is reasonable to think that it will continue to be used and further refined as new data types arise (e.g. single cell datasets) and the identification of system-level patterns will be considered more and more important in omics data analysis.

Pathogenic mutations and overall survival in 3,084 patients with cancer: the Hellenic Cooperative Oncology Group Precision Medicine Initiative

Background: We evaluated the association between pathogenic mutations and overall survival (OS) in patients with cancer referred to Hellenic Cooperative Oncology Group–affiliated Departments.

Patients and methods: Patients referred from 12/1980 to 1/2017 had molecular testing (for research) of archival tumor tissue collected at the time of first diagnosis (non-metastatic, 81%; metastatic, 19%). Tumor-specific gene panels (16-101 genes) were used to identify pathogenic mutations in clinically relevant genes. NGS genotyping was performed at the Laboratory of Molecular Oncology, Aristotle University of Thessaloniki. Annotation of mutations was performed at MD Anderson Cancer Center.

Results: We analyzed 3,084 patients (median age, 57 years; men, 22%) with sequencing data. Overall, 1,775 (58% of 3,084) patients had pathogenic mutations. The median follow-up was 7.52 years (95% CI, 7.39-7.61). In patients with non-metastatic tumors, after stratification by tumor type, increasing age, higher grade, and histology other than adenocarcinoma were associated with shorter OS. OS was also shorter in patients with pathogenic TP53 (HR=1.36; p<0.001), MLL3 (HR=1.64; p=0.005), and BRCA1 (HR=1.46; p=0.047) mutations compared to wild-type genes. In multivariate analyses, independent prognostic factors predicting shorter OS were pathogenic mutations in TP53 (HR=1.37, p=0.002) and MLL3 (HR=1.50, p=0.027); increasing age (HR=1.02, p<0.001); and increasing grade (HR=1.46, p<0.001). In patients with metastatic cancer, older age and higher grade were associated with shorter OS and maintained their independent prognostic significance (increasing age, HR=1.03, p<0.001 and higher grade, HR=1.73, p<0.001).

Conclusions: Analysis of molecular data reveals prognostic biomarkers, regardless of tissue or organ of origin to improve patient management.

Predicting links between tumor samples and genes using 2-Layered graph based diffusion approach

Background

Determining the association between tumor sample and the gene is demanding because it requires a high cost for conducting genetic experiments. Thus, the discovered association between tumor sample and gene further requires clinical verification and validation. This entire mechanism is time-consuming and expensive. Due to this issue, predicting the association between tumor samples and genes remain a challenge in biomedicine.

Results

Here we present, a computational model based on a heat diffusion algorithm which can predict the association between tumor samples and genes. We proposed a 2-layered graph. In the first layer, we constructed a graph of tumor samples and genes where these two types of nodes are connected by “hasGene” relationship. In the second layer, the gene nodes are connected by “interaction” relationship. We applied the heat diffusion algorithms in nine different variants of genetic interaction networks extracted from STRING and BioGRID database. The heat diffusion algorithm predicted the links between tumor samples and genes with mean AUC-ROC score of 0.84. This score is obtained by using weighted genetic interactions of fusion or co-occurrence channels from the STRING database. For the unweighted genetic interaction from the BioGRID database, the algorithms predict the links with an AUC-ROC score of 0.74.

Conclusions

We demonstrate that the gene-gene interaction scores could improve the predictive power of the heat diffusion model to predict the links between tumor samples and genes. We showed the efficient runtime of the heat diffusion algorithm in various genetic interaction network. We statistically validated our prediction quality of the links between tumor samples and genes.

Electronic supplementary material

The online version of this article (10.1186/s12859-019-3056-2) contains supplementary material, which is available to authorized users.

Tissue specificity of in vitro drug sensitivity

Objectives

We sought to investigate the tissue specificity of drug sensitivities in large-scale pharmacological studies and compare these associations to those found in drug clinical indications.

Materials and Methods

We leveraged the curated cell line response data from PharmacoGx and applied an enrichment algorithm on drug sensitivity values' area under the drug dose-response curves (AUCs) with and without adjustment for general level of drug sensitivity.

Results

We observed tissue specificity in 63% of tested drugs, with 8% of total interactions deemed significant (false discovery rate <0.05). By restricting the drug-tissue interactions to those with AUC > 0.2, we found that in 52% of interactions, the tissue was predictive of drug sensitivity (concordance index > 0.65). When compared with clinical indications, the observed overlap was weak (Matthew correlation coefficient, MCC = 0.0003, P > .10).

Discussion

While drugs exhibit significant tissue specificity in vitro, there is little overlap with clinical indications. This can be attributed to factors such as underlying biological differences between in vitro models and patient tumors, or the inability of tissue-specific drugs to bring additional benefits beyond gold standard treatments during clinical trials.

Conclusion

Our meta-analysis of pan-cancer drug screening datasets indicates that most tested drugs exhibit tissue-specific sensitivities in a large panel of cancer cell lines. However, the observed preclinical results do not translate to the clinical setting. Our results suggest that additional research into showing parallels between preclinical and clinical data is required to increase the translational potential of in vitro drug screening.

Computational Methods for Subtyping of Tumors and Their Applications for Deciphering Tumor Heterogeneity

With the rapid development of deep sequencing technologies, many programs are generating multi-platform genomic profiles (e.g., somatic mutation, DNA methylation, and gene expression) for a large number of tumors. This activity has provided unique opportunities and challenges to stratify tumors and decipher tumor heterogeneity. In this chapter, we summarize several computational methods to address the challenge of tumor stratification with different types of genomic data. We further introduce their applications in emerging large-scale genomic data to show their effectiveness in deciphering tumor heterogeneity and clinical relevance.

Dissecting the expression landscape of mitochondrial genes in lung squamous cell carcinoma and lung adenocarcinoma

Lung squamous cell carcinoma (LUSC) and lung adenocarcinoma (LUAD) are the two major subtypes of lung cancer. To explore mitochondrial respiratory gene expression profiles in LUSC and LUAD, RNA sequencing data from The Cancer Genome Atlas was used for comprehensive analyses to establish the molecular characteristics of LUSC and LUAD. To elucidate expression profiles, subtypes were defined using unsupervised clustering of mitochondrial gene expression data. Differences in nuclear gene expression levels, signaling pathways and tumor microenvironments between subtypes were investigated. The analysis revealed that mitochondrial respiratory genes were generally expressed at lower levels in tumor tissues compared with matched control tissues. The expression of mitochondrially encoded NADH dehydrogenase 5 or 6 was associated with tumor progression in LUAD and LUSC. Patients were clustered into three subgroups based on the expression profile of 13 mitochondrial protein-encoding genes, and patients in Cluster 3 exhibited poor survival rates compared with patients from Cluster 1. Furthermore, this association was also observed in another independent data set. Further analyses of the expression of nuclear-encoded genes in the three clusters revealed the enrichment of several cancer-associated signaling pathways in Cluster 3, particularly the apoptotic signaling pathway, suggesting a potential association between the decreased expression of mitochondrial DNA genes and increased tumor aggressiveness. Furthermore, the analyses of immune cell compositions in the tumor microenvironment detected a significant increase in the proportion of CD4+ T cells and a decrease in the proportion of macrophages in LUAD compared with LUSC (P=0.0000104 and P=0.0000105, respectively). In conclusion, the present study revealed an association between the expression patterns of mitochondrial-encoded genes and lung cancer, which may contribute to novel therapeutic strategies for patients with LUSC and LUAD.

The Discovery of Mutated Driver Pathways in Cancer: Models and Algorithms

The pathogenesis of cancer in human is still poorly understood. With the rapid development of high-throughput sequencing technologies, huge volumes of cancer genomics data have been generated. Deciphering that data poses great opportunities and challenges to computational biologists. One of such key challenges is to distinguish driver mutations, genes as well as pathways from passenger ones. Mutual exclusivity of gene mutations (each patient has no more than one mutation in the gene set) has been observed in various cancer types and thus has been used as an important property of a driver gene set or pathway. In this article, we aim to review the recent development of computational models and algorithms for discovering driver pathways or modules in cancer with the focus on mutual exclusivity-based ones.

Pan-cancer analysis of somatic mutations and transcriptomes reveals common functional gene clusters shared by multiple cancer types

To discover functional gene clusters across cancers, we performed a systematic pan-cancer analysis of 33 cancer types. We identified genes that were associated with somatic mutations and were the cores of a co-expression network. We found that multiple cancer types have relatively exclusive hub genes individually; however, the hub genes cooperate with each other based on their functional relationship. When we built a protein-protein interaction network of hub genes and found nine functional gene clusters across cancer types, the gene clusters divided not only the region of the network map, but also the function of the network by their distinct roles related to the development and progression of cancer. This functional relationship between the clusters and cancers was underpinned by the high expression of module genes and enrichment of programmed cell death, and known candidate cancer genes. In addition to protein-coding hub genes, non-coding hub genes had a possible relationship with cancer. Overall, our approach of investigating cancer genes enabled finding pan-cancer hub genes and common functional gene clusters shared by multiple cancer types based on the expression status of the primary tumour and the functional relationship of genes in the biological network.

WebPropagate: A Web Server for Network Propagation

Network propagation is a powerful tool for genetic analysis which is widely used to identify genes and genetic modules that underlie a process of interest. Here we provide a graphical, web-based platform (http://anat.cs.tau.ac.il/WebPropagate/) in which researchers can easily apply variants of this method to data sets of interest using up-to-date networks of protein-protein interactions in several organisms.

NSD1 inactivation defines an immune cold, DNA hypomethylated subtype in squamous cell carcinoma

Chromatin modifying enzymes are frequently mutated in cancer, resulting in widespread epigenetic deregulation. Recent reports indicate that inactivating mutations in the histone methyltransferase NSD1 define an intrinsic subtype of head and neck squamous cell carcinoma (HNSC) that features pronounced DNA hypomethylation. Here, we describe a similar hypomethylated subtype of lung squamous cell carcinoma (LUSC) that is enriched for both inactivating mutations and deletions in NSD1. The ‘NSD1 subtypes’ of HNSC and LUSC are highly correlated at the DNA methylation and gene expression levels, featuring ectopic expression of developmental transcription factors and genes that are also hypomethylated in Sotos syndrome, a congenital disorder caused by germline NSD1 mutations. Further, the NSD1 subtype of HNSC displays an ‘immune cold’ phenotype characterized by low infiltration of tumor-associated leukocytes, particularly macrophages and CD8⁺ T cells, as well as low expression of genes encoding the immunotherapy target PD-1 immune checkpoint receptor and its ligands. Using an in vivo model, we demonstrate that NSD1 inactivation results in reduced T cell infiltration into the tumor microenvironment, implicating NSD1 as a tumor cell-intrinsic driver of an immune cold phenotype. NSD1 inactivation therefore causes epigenetic deregulation across cancer sites, and has implications for immunotherapy.

Network propagation in the cytoscape cyberinfrastructure

Network propagation is an important and widely used algorithm in systems biology, with applications in protein function prediction, disease gene prioritization, and patient stratification. However, up to this point it has required significant expertise to run. Here we extend the popular network analysis program Cytoscape to perform network propagation as an integrated function. Such integration greatly increases the access to network propagation by putting it in the hands of biologists and linking it to the many other types of network analysis and visualization available through Cytoscape. We demonstrate the power and utility of the algorithm by identifying mutations conferring resistance to Vemurafenib.

Discovery of cancer common and specific driver gene sets

Cancer is known as a disease mainly caused by gene alterations. Discovery of mutated driver pathways or gene sets is becoming an important step to understand molecular mechanisms of carcinogenesis. However, systematically investigating commonalities and specificities of driver gene sets among multiple cancer types is still a great challenge, but this investigation will undoubtedly benefit deciphering cancers and will be helpful for personalized therapy and precision medicine in cancer treatment. In this study, we propose two optimization models to de novo discover common driver gene sets among multiple cancer types (ComMDP) and specific driver gene sets of one certain or multiple cancer types to other cancers (SpeMDP), respectively. We first apply ComMDP and SpeMDP to simulated data to validate their efficiency. Then, we further apply these methods to 12 cancer types from The Cancer Genome Atlas (TCGA) and obtain several biologically meaningful driver pathways. As examples, we construct a common cancer pathway model for BRCA and OV, infer a complex driver pathway model for BRCA carcinogenesis based on common driver gene sets of BRCA with eight cancer types, and investigate specific driver pathways of the liquid cancer lymphoblastic acute myeloid leukemia (LAML) versus other solid cancer types. In these processes more candidate cancer genes are also found.

Network based stratification of major cancers by integrating somatic mutation and gene expression data

The stratification of cancer into subtypes that are significantly associated with clinical outcomes is beneficial for targeted prognosis and treatment. In this study, we integrated somatic mutation and gene expression data to identify clusters of patients. In contrast to previous studies, we constructed cancer-type-specific significant co-expression networks (SCNs) rather than using a fixed gene network across all cancers, such as the network-based stratification (NBS) method, which ignores cancer heterogeneity. For each type of cancer, the gene expression data were used to construct the SCN network, while the gene somatic mutation data were mapped onto the network, propagated, and used for further clustering. For the clustering, we adopted an improved network-regularized non-negative matrix factorization (netNMF) (netNMF_HC) for a more precise classification. We applied our method to various datasets, including ovarian cancer (OV), lung adenocarcinoma (LUAD) and uterine corpus endometrial carcinoma (UCEC) cohorts derived from the TCGA (The Cancer Genome Atlas) project. Based on the results, we evaluated the performance of our method to identify survival-relevant subtypes and further compared it to the NBS method, which adopts priori networks and netNMF algorithm. The proposed algorithm outperformed the NBS method in identifying informative cancer subtypes that were significantly associated with clinical outcomes in most cancer types we studied. In particular, our method identified survival-associated UCEC subtypes that were not identified by the NBS method. Our analysis indicated valid subtyping of patient could be applied by mutation data with cancer-type-specific SCNs and netNMF_HC for individual cancers because of specific cancer co-expression patterns and more precise clustering.

Trans-acting epigenetic effects of chromosomal aneuploidies: lessons from Down syndrome and mouse models

An important line of postgenomic research seeks to understand how genetic factors can influence epigenetic patterning. Here we review epigenetic effects of chromosomal aneuploidies, focusing on findings in Down syndrome (DS, trisomy 21). Recent work in human DS and mouse models has shown that the extra chromosome 21 acts in trans to produce epigenetic changes, including differential CpG methylation (DS-DM), in specific sets of downstream target genes, mostly on other chromosomes. Mechanistic hypotheses emerging from these data include roles of chromosome 21-linked methylation pathway genes (DNMT3L and others) and transcription factor genes (RUNX1, OLIG2, GABPA, ERG and ETS2) in shaping the patterns of DS-DM. The findings may have broader implications for trans-acting epigenetic effects of chromosomal and subchromosomal aneuploidies in other human developmental and neuropsychiatric disorders, and in cancers.

An integrative and comparative study of pan-cancer transcriptomes reveals distinct cancer common and specific signatures

To investigate the commonalities and specificities across tumor lineages, we perform a systematic pan-cancer transcriptomic study across 6744 specimens. We find six pan-cancer subnetwork signatures which relate to cell cycle, immune response, Sp1 regulation, collagen, muscle system and angiogenesis. Moreover, four pan-cancer subnetwork signatures demonstrate strong prognostic potential. We also characterize 16 cancer type-specific subnetwork signatures which show diverse implications to somatic mutations, somatic copy number aberrations, DNA methylation alterations and clinical outcomes. Furthermore, some of them are strongly correlated with histological or molecular subtypes, indicating their implications with tumor heterogeneity. In summary, we systematically explore the pan-cancer common and cancer type-specific gene subnetwork signatures across multiple cancers, and reveal distinct commonalities and specificities among cancers at transcriptomic level.

A microscopic landscape of the invasive breast cancer genome

Histologic grade is one of the most important microscopic features used to predict the prognosis of invasive breast cancer and may serve as a marker for studying cancer driving genomic abnormalities in vivo. We analyzed whole genome sequencing data from 680 cases of TCGA invasive ductal carcinomas of the breast and correlated them to corresponding pathology information. Ten genetic abnormalities were found to be statistically associated with histologic grade, including three most prevalent cancer driver events, TP53 and PIK3CA mutations and MYC amplification. A distinct genetic interaction among these genomic abnormalities was revealed as measured by the histologic grading score. While TP53 mutation and MYC amplification were synergistic in promoting tumor progression, PIK3CA mutation was found to have alleviated the oncogenic effect of either the TP53 mutation or MYC amplification, and was associated with a significant reduction in mitotic activity in TP53 mutated and/or MYC amplified breast cancer. Furthermore, we discovered that different types of genetic abnormalities (mutation versus amplification) within the same cancer driver gene (PIK3CA or GATA3) were associated with opposite histologic changes in invasive breast cancer. In conclusion, our study suggests that histologic grade may serve as a biomarker to define cancer driving genetic events in vivo.

Fast dimension reduction and integrative clustering of multi-omics data using low-rank approximation: application to cancer molecular classification

Background

One major goal of large-scale cancer omics study is to identify molecular subtypes for more accurate cancer diagnoses and treatments. To deal with high-dimensional cancer multi-omics data, a promising strategy is to find an effective low-dimensional subspace of the original data and then cluster cancer samples in the reduced subspace. However, due to data-type diversity and big data volume, few methods can integrative and efficiently find the principal low-dimensional manifold of the high-dimensional cancer multi-omics data.

Results

In this study, we proposed a novel low-rank approximation based integrative probabilistic model to fast find the shared principal subspace across multiple data types: the convexity of the low-rank regularized likelihood function of the probabilistic model ensures efficient and stable model fitting. Candidate molecular subtypes can be identified by unsupervised clustering hundreds of cancer samples in the reduced low-dimensional subspace. On testing datasets, our method LRAcluster (low-rank approximation based multi-omics data clustering) runs much faster with better clustering performances than the existing method. Then, we applied LRAcluster on large-scale cancer multi-omics data from TCGA. The pan-cancer analysis results show that the cancers of different tissue origins are generally grouped as independent clusters, except squamous-like carcinomas. While the single cancer type analysis suggests that the omics data have different subtyping abilities for different cancer types.

Conclusions

LRAcluster is a very useful method for fast dimension reduction and unsupervised clustering of large-scale multi-omics data. LRAcluster is implemented in R and freely available via http://bioinfo.au.tsinghua.edu.cn/software/lracluster/.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-2223-8) contains supplementary material, which is available to authorized users.