Identifying cancer genes from cancer mutation profiles by cancer functions

It is of great importance to identify new cancer genes from the data of large scale genome screenings of gene mutations in cancers. Considering the alternations of some essential functions are indispensable for oncogenesis, we define them as cancer functions and select, as their approximations, a group of detailed functions in GO (Gene Ontology) highly enriched with known cancer genes. To evaluate the efficiency of using cancer functions as features to identify cancer genes, we define, in the screened genes, the known protein kinase cancer genes as gold standard positives and the other kinase genes as gold standard negatives. The results show that cancer associated functions are more efficient in identifying cancer genes than the selection pressure feature. Furthermore, combining cancer functions with the number of non-silent mutations can generate more reliable positive predictions. Finally, with precision 0.42, we suggest a list of 46 kinase genes as candidate cancer genes which are annotated to cancer functions and carry at least 3 non-silent mutations.     

Evolutionary dynamics of oncogenes and tumor suppressor genes: higher intensities of purifying selection than other genes

Oncogenes and tumor suppressor genes (hereafter referred to as "cancer genes") result in cancer when they experience substitutions that prevent or distort their normal function. We examined evolutionary pressures acting on cancer genes and other classes of disease-related genes and compared our results to analyses of genes without known association to disease. We compared synonymous and nonsynonymous substitution rates in 3,035 human genes-approximately 10% of the genome-measuring the intensity of purifying selection on 311 human disease genes, including 122 cancer-related genes. Although the genes examined are similar to nondisease genes in product, expression, function, and pathway affiliation, we found intriguing differences in the selective pressures experienced by cancer genes relative to other (noncancer) disease-related and non-disease-related genes. We found a statistically significant increase in the intensity of purifying selection exerted on cancer genes (the average ratio of nonsynonymous to synonymous substitutions, omega, was 0.079) relative to all other disease-related genes groups (omega = 0.101) and non-disease-related genes (omega = 0.100). This difference indicates a striking increase in selection against nonsynonymous substitutions in oncogenes and tumor suppressor genes. This finding provides insight into the etiology of cancer and the differences between genes involved in cancer and those implicated in other human diseases. Specifically, we found a significant overlap between human oncogenes and tumor suppressor genes and "essential genes," human homologs of mouse lethal genes identified by knockout experiments. This insight may improve our ability to identify cancer-related genes and enhances our understanding of the nature of these genes.     

Prioritization of candidate cancer genes--an aid to oncogenomic studies

The development of techniques for oncogenomic analyses such as array comparative genomic hybridization, messenger RNA expression arrays and mutational screens have come to the fore in modern cancer research. Studies utilizing these techniques are able to highlight panels of genes that are altered in cancer. However, these candidate cancer genes must then be scrutinized to reveal whether they contribute to oncogenesis or are coincidental and non-causative. We present a computational method for the prioritization of candidate (i) proto-oncogenes and (ii) tumour suppressor genes from oncogenomic experiments. We constructed computational classifiers using different combinations of sequence and functional data including sequence conservation, protein domains and interactions, and regulatory data. We found that these classifiers are able to distinguish between known cancer genes and other human genes. Furthermore, the classifiers also discriminate candidate cancer genes from a recent mutational screen from other human genes. We provide a web-based facility through which cancer biologists may access our results and we propose computational cancer gene classification as a useful method of prioritizing candidate cancer genes identified in oncogenomic studies.     

根据在癌突变谱中的共突变基因对识别癌基因

癌是一种涉及多基因变异的遗传异质性疾病,涉及多种生物学功能通路中不同基因的遗传变异。因此,识别癌基因是一项富有挑战性的工作。提出通过寻找在癌样本中突变显著共发生的基因筛选候选癌基因的方法。应用该方法,通过分析蛋白激酶基因在癌组织中的突变谱数据,发现了167个显著共发生突变的基因对,包含85个基因。分析这167个基因对发现:(1)发生共突变的基因富集已知的癌基因;(2)共突变基因对倾向于共扰动与癌症相关的通路对。以上结果提示,在癌样本中显著共发生突变的基因倾向于候选癌基因;在癌发生过程中起重要作用的基因倾向于协同扰动不同的癌相关细胞生物学过程。     

MSEA: detection and quantification of mutation hotspots through mutation set enrichment analysis

Many cancer genes form mutation hotspots that disrupt their functional domains or active sites, leading to gain- or loss-of-function. We propose a mutation set enrichment analysis (MSEA) implemented by two novel methods, MSEA-clust and MSEA-domain, to predict cancer genes based on mutation hotspot patterns. MSEA methods are evaluated by both simulated and real cancer data. We find approximately 51% of the eligible known cancer genes form detectable mutation hotspots. Application of MSEA in eight cancers reveals a total of 82 genes with mutation hotspots, including well-studied cancer genes, known cancer genes re-found in new cancer types, and novel cancer genes.

Electronic supplementary material

The online version of this article (doi:10.1186/s13059-014-0489-9) contains supplementary material, which is available to authorized users.     

Analysis of gene expression profiles of gastric normal and cancer tissues by SAGE

In an attempt to understand the molecular bases of gastric cancer progression, we have analyzed the differentially expressed genes in gastric cancer by SAGE. Four SAGE cDNA tag libraries were constructed from two sets of gastric cancer and normal tissues and 241,127 tags were obtained. By comparing the tags from cancer and normal tissues, 414 differentially expressed tags, representing 383 genes, were identified in cancer tissues (p 相似文献   

Overview: genes that predispose to cancer

Heredity and environment both operate in the origin of cancer. Dominantly heritable cancer is caused by 'cancer' genes that impart high relative risks but account for only a small part of the incidence of cancer; they are usually recessive in oncogenesis, mutation or loss of the second allele being necessary. Non-hereditary forms of cancer may involve the same genes. Other genes interact with environment in carcinogenesis; these may impart relatively small relative risks, but because their frequencies may be high, the attributable risks can be great, as probably is the case with lung cancer. The process of carcinogenesis is thought to involve 2 or more somatic genetic events in most cases. The genes whose germline mutations cause dominantly inherited cancer can also be mutated somatically to cause non-hereditary cancer. Other genes may influence the numbers of target cells, or the proliferation of once-hit stem cells, without being critical events on the path to cancer. However, such genes could greatly influence the incidence of a cancer. Other genes, such as that for Bloom's syndrome, may affect the rates at which first and second events occur. Finally, other genes may influence the occurrence of events critical for progression and metastasis, such as vascularization of a small tumor.     

Comparison of expression profiles in ovarian epithelium in vivo and ovarian cancer identifies novel candidate genes involved in disease pathogenesis

Molecular events leading to epithelial ovarian cancer are poorly understood but ovulatory hormones and a high number of life-time ovulations with concomitant proliferation, apoptosis, and inflammation, increases risk. We identified genes that are regulated during the estrous cycle in murine ovarian surface epithelium and analysed these profiles to identify genes dysregulated in human ovarian cancer, using publically available datasets. We identified 338 genes that are regulated in murine ovarian surface epithelium during the estrous cycle and dysregulated in ovarian cancer. Six of seven candidates selected for immunohistochemical validation were expressed in serous ovarian cancer, inclusion cysts, ovarian surface epithelium and in fallopian tube epithelium. Most were overexpressed in ovarian cancer compared with ovarian surface epithelium and/or inclusion cysts (EpCAM, EZH2, BIRC5) although BIRC5 and EZH2 were expressed as highly in fallopian tube epithelium as in ovarian cancer. We prioritised the 338 genes for those likely to be important for ovarian cancer development by in silico analyses of copy number aberration and mutation using publically available datasets and identified genes with established roles in ovarian cancer as well as novel genes for which we have evidence for involvement in ovarian cancer. Chromosome segregation emerged as an important process in which genes from our list of 338 were over-represented including two (BUB1, NCAPD2) for which there is evidence of amplification and mutation. NUAK2, upregulated in ovarian surface epithelium in proestrus and predicted to have a driver mutation in ovarian cancer, was examined in a larger cohort of serous ovarian cancer where patients with lower NUAK2 expression had shorter overall survival. In conclusion, defining genes that are activated in normal epithelium in the course of ovulation that are also dysregulated in cancer has identified a number of pathways and novel candidate genes that may contribute to the development of ovarian cancer.     

A novel algorithm based on bi-random walks to identify disease-related lncRNAs

Background

It’s a very urgent task to identify cancer genes that enables us to understand the mechanisms of biochemical processes at a biomolecular level and facilitates the development of bioinformatics. Although a large number of methods have been proposed to identify cancer genes at recent times, the biological data utilized by most of these methods is still quite less, which reflects an insufficient consideration of the relationship between genes and diseases from a variety of factors.

Results

In this paper, we propose a two-rounds random walk algorithm to identify cancer genes based on multiple biological data (TRWR-MB), including protein-protein interaction (PPI) network, pathway network, microRNA similarity network, lncRNA similarity network, cancer similarity network and protein complexes. In the first-round random walk, all cancer nodes, cancer-related genes, cancer-related microRNAs and cancer-related lncRNAs, being associated with all the cancer, are used as seed nodes, and then a random walker walks on a quadruple layer heterogeneous network constructed by multiple biological data. The first-round random walk aims to select the top score k of potential cancer genes. Then in the second-round random walk, genes, microRNAs and lncRNAs, being associated with a certain special cancer in corresponding cancer class, are regarded as seed nodes, and then the walker walks on a new quadruple layer heterogeneous network constructed by lncRNAs, microRNAs, cancer and selected potential cancer genes. After the above walks finish, we combine the results of two-rounds RWR as ranking score for experimental analysis. As a result, a higher value of area under the receiver operating characteristic curve (AUC) is obtained. Besides, cases studies for identifying new cancer genes are performed in corresponding section.

Conclusion

In summary, TRWR-MB integrates multiple biological data to identify cancer genes by analyzing the relationship between genes and cancer from a variety of biological molecular perspective.

     

Genome wide identification of recessive cancer genes by combinatorial mutation analysis

We devised a novel procedure to identify human cancer genes acting in a recessive manner. Our strategy was to combine the contributions of the different types of genetic alterations to loss of function: amino-acid substitutions, frame-shifts, gene deletions. We studied over 20,000 genes in 3 Gigabases of coding sequences and 700 array comparative genomic hybridizations. Recessive genes were scored according to nucleotide mismatches under positive selective pressure, frame-shifts and genomic deletions in cancer. Four different tests were combined together yielding a cancer recessive p-value for each studied gene. One hundred and fifty four candidate recessive cancer genes (p-value < 1.5 x 10(-7), FDR = 0.39) were identified. Strikingly, the prototypical cancer recessive genes TP53, PTEN and CDKN2A all ranked in the top 0.5% genes. The functions significantly affected by cancer mutations are exactly overlapping those of known cancer genes, with the critical exception for the absence of tyrosine kinases, as expected for a recessive gene-set.     

Network-based integration method for potential breast cancer gene identification

Breast cancer is the most common female death-causing cancer worldwide. A network-based integration method was proposed to identify potential breast cancer genes. First, genes were prioritized using a gene prioritization algorithm by the strategy of disease risks transferred between genes in a network with weighted vertexes and edges. Our prioritization algorithm was effectives and robust for top-ranked seed gene number and higher area under the curve values compared to ToppGene and ToppNet. Then, 20 potential breast cancer genes were identified as common genes of the top 50 candidate genes for their robustness in multiple prioritizations. These genes could accurately classify tumor and normal samples of all and paired sample sets and three independent datasets. Of potential breast cancer genes, 18 were verified by literature and 2 were novel genes that need further study. This study would contribute to the understanding of the genetic architecture for the diagnosis and treatment of breast cancer.     

A Network of Cancer Genes with Co-Occurring and Anti-Co-Occurring Mutations

Certain cancer genes contribute to tumorigenesis in a manner of either co-occurring or mutually exclusive (anti-co-occurring) mutations; however, the global picture of when, where and how these functional interactions occur remains unclear. This study presents a systems biology approach for this purpose. After applying this method to cancer gene mutation data generated from large-scale and whole genome sequencing of cancer samples, a network of cancer genes with co-occurring and anti-co-occurring mutations was constructed. Analysis of this network revealed that genes with co-occurring mutations prefer direct signaling transductions and that the interaction relations among cancer genes in the network are related with their functional similarity. It was also revealed that genes with co-occurring mutations tend to have similar mutation frequencies, whereas genes with anti-co-occurring mutations tend to have different mutation frequencies. Moreover, genes with more exons tend to have more co-occurring mutations with other genes, and genes having lower local coherent network structures tend to have higher mutation frequency. The network showed two complementary modules that have distinct functions and have different roles in tumorigenesis. This study presented a framework for the analysis of cancer genome sequencing outputs. The presented data and uncovered patterns are helpful for understanding the contribution of gene mutations to tumorigenesis and valuable in the identification of key biomarkers and drug targets for cancer.     

Breast cancer gene discovery

Many important advances have been made in the past decade in understanding breast cancer at the molecular level, and two important high-penetrance breast cancer genes--BRCA1 and BRCA2--have been identified. However, germline mutations in these two genes are responsible for only a minority (approximately 5%) of all breast carcinomas, and the genes responsible for the majority of breast cancer cases remain to be identified. There is evidence that there are additional high-to-moderate-penetrance breast cancer susceptibility genes but, given the high degree of molecular heterogeneity in breast carcinomas, it is likely that each of these genes is responsible for only a subset of cases. There are also many candidate low-penetrance breast cancer genes and many more are likely to be identified. In addition to germline, and somatic, sequence alterations, epigenetic changes in many genes are likely to play an important role in the pathobiology of breast cancer. Recently developed genomic technologies and the completion of the human genome sequence provide us with powerful tools to identify novel candidate breast cancer genes that could play an important role in breast tumourigenesis.     

How germline genes promote malignancy in cancer cells

Cancers often express hundreds of genes otherwise specific to germ cells, the germline/cancer (GC) genes. Here, we present and discuss the hypothesis that activation of a “germline program” promotes cancer cell malignancy. We do so by proposing four hallmark processes of the germline: meiosis, epigenetic plasticity, migration, and metabolic plasticity. Together, these hallmarks enable replicative immortality of germ cells as well as cancer cells. Especially meiotic genes are frequently expressed in cancer, implying that genes unique to meiosis may play a role in oncogenesis. Because GC genes are not expressed in healthy somatic tissues, they form an appealing source of specific treatment targets with limited side effects besides infertility. Although it is still unclear why germ cell specific genes are so abundantly expressed in cancer, from our hypothesis it follows that the germline's reproductive program is intrinsic to cancer development.     

Cancer Evolution Is Associated with Pervasive Positive Selection on Globally Expressed Genes

Cancer is an evolutionary process in which cells acquire new transformative, proliferative and metastatic capabilities. A full understanding of cancer requires learning the dynamics of the cancer evolutionary process. We present here a large-scale analysis of the dynamics of this evolutionary process within tumors, with a focus on breast cancer. We show that the cancer evolutionary process differs greatly from organismal (germline) evolution. Organismal evolution is dominated by purifying selection (that removes mutations that are harmful to fitness). In contrast, in the cancer evolutionary process the dominance of purifying selection is much reduced, allowing for a much easier detection of the signals of positive selection (adaptation). We further show that, as a group, genes that are globally expressed across human tissues show a very strong signal of positive selection within tumors. Indeed, known cancer genes are enriched for global expression patterns. Yet, positive selection is prevalent even on globally expressed genes that have not yet been associated with cancer, suggesting that globally expressed genes are enriched for yet undiscovered cancer related functions. We find that the increased positive selection on globally expressed genes within tumors is not due to their expression in the tissue relevant to the cancer. Rather, such increased adaptation is likely due to globally expressed genes being enriched in important housekeeping and essential functions. Thus, our results suggest that tumor adaptation is most often mediated through somatic changes to those genes that are important for the most basic cellular functions. Together, our analysis reveals the uniqueness of the cancer evolutionary process and the particular importance of globally expressed genes in driving cancer initiation and progression.     

Predicting potential cancer genes by integrating network properties, sequence features and functional annotations

The discovery of novel cancer genes is one of the main goals in cancer research. Bioinformatics methods can be used to accelerate cancer gene discovery, which may help in the understanding of cancer and the development of drug targets. In this paper, we describe a classifier to predict potential cancer genes that we have developed by integrating multiple biological evidence, including protein-protein interaction network properties, and sequence and functional features. We detected 55 features that were significantly different between cancer genes and non-cancer genes. Fourteen cancer-associated features were chosen to train the classifier. Four machine learning methods, logistic regression, support vector machines (SVMs), BayesNet and decision tree, were explored in the classifier models to distinguish cancer genes from non-cancer genes. The prediction power of the different models was evaluated by 5-fold cross-validation. The area under the receiver operating characteristic curve for logistic regression, SVM, Baysnet and J48 tree models was 0.834, 0.740, 0.800 and 0.782, respectively. Finally, the logistic regression classifier with multiple biological features was applied to the genes in the Entrez database, and 1976 cancer gene candidates were identified. We found that the integrated prediction model performed much better than the models based on the individual biological evidence, and the network and functional features had stronger powers than the sequence features in predicting cancer genes.     

Complexities of cancer research: mouse genetic models

Cancer susceptibility is a complex interaction of an individual's genetic composition and environmental exposures. Huge strides have been made in understanding cancer over the past 100 yr, from recognition of cancer as a genetic disease, to identification of specific carcinogens, isolation of oncogenes, and recognition of tumor suppressors. A tremendous amount of knowledge has accumulated about the etiology of cancer. Cancer genetics has played a significant role in these discoveries. Analysis of high-risk familial cancers has led to the discovery of new tumor suppressor genes and important cancer pathways. These families, however, represent only a small fraction of cancer in the general population. Most cancer is instead probably the result of an intricate interaction of polymorphic susceptibility genes with the sea of environmental exposures that humans experience. Although the central cadre of cancer genes is known, little is understood about the peripheral genes that likely comprise the polymorphic susceptibility loci. The challenge for cancer genetics is therefore to move forward from the mendelian genetics of the rare familial cancer syndromes into the field of quantitative trait loci, susceptibility factors, and modifier genes. By identifying the genes that modulate an individual's susceptibility to cancer after an environmental exposure, researchers will be able to gain important insights into human biology, cancer prevention, and cancer treatment. This article summarizes the current state of quantitative trait genetic analysis and the tools, both proven and theoretical, that may be used to unravel one of the great challenges in cancer genetics.