Tiziana Sanavia

Effect of Size and Heterogeneity of Samples on Biomarker Discovery: Synthetic and Real Data Assessment

Motivation The identification of robust lists of molecular biomarkers related to a disease is a fundamental step for early diagnosis and treatment. However, methodologies for the discovery of biomarkers using microarray data often provide results with limited overlap. These differences are imputable to 1) dataset size (few subjects with respect to the number of features); 2) heterogeneity of the disease; 3) heterogeneity of experimental protocols and computational pipelines employed in the analysis. In this paper, we focus on the first two issues and assess, both on simulated (through an in silico regulation network model) and real clinical datasets, the consistency of candidate biomarkers provided by a number of different methods. Methods We extensively simulated the effect of heterogeneity characteristic of complex diseases on different sets of microarray data. Heterogeneity was reproduced by simulating both intrinsic variability of the population and the alteration of regulatory mechanisms. Population variability was simulated by modeling evolution of a pool of subjects; then, a subset of them underwent alterations in regulatory mechanisms so as to mimic the disease state. Results The simulated data allowed us to outline advantages and drawbacks of different methods across multiple studies and varying number of samples and to evaluate precision of feature selection on a benchmark with known biomarkers. Although comparable classification accuracy was reached by different methods, the use of external cross-validation loops is helpful in finding features with a higher degree of precision and stability. Application to real data confirmed these results.

FunPat: function-based pattern analysis on RNA-seq time series data

BackgroundDynamic expression data, nowadays obtained using high-throughput RNA sequencing, are essential to monitor transient gene expression changes and to study the dynamics of their transcriptional activity in the cell or response to stimuli. Several methods for data selection, clustering and functional analysis are available; however, these steps are usually performed independently, without exploiting and integrating the information derived from each step of the analysis.MethodsHere we present FunPat, an R package for time series RNA sequencing data that integrates gene selection, clustering and functional annotation into a single framework. FunPat exploits functional annotations by performing for each functional term, e.g. a Gene Ontology term, an integrated selection-clustering analysis to select differentially expressed genes that share, besides annotation, a common dynamic expression profile.ResultsFunPat performance was assessed on both simulated and real data. With respect to a stand-alone selection step, the integration of the clustering step is able to improve the recall without altering the false discovery rate. FunPat also shows high precision and recall in detecting the correct temporal expression patterns; in particular, the recall is significantly higher than hierarchical, k-means and a model-based clustering approach specifically designed for RNA sequencing data. Moreover, when biological replicates are missing, FunPat is able to provide reproducible lists of significant genes. The application to real time series expression data shows the ability of FunPat to select differentially expressed genes with high reproducibility, indirectly confirming high precision and recall in gene selection. Moreover, the expression patterns obtained as output allow an easy interpretation of the results.ConclusionsA novel analysis pipeline was developed to search the main temporal patterns in classes of genes similarly annotated, improving the sensitivity of gene selection by integrating the statistical evidence of differential expression with the information on temporal profiles and the functional annotations. Significant genes are associated to both the most informative functional terms, avoiding redundancy of information, and the most representative temporal patterns, thus improving the readability of the results. FunPat package is provided in R/Bioconductor at link: http://sysbiobig.dei.unipd.it/?q=node/79.

Improving biomarker list stability by integration of biological knowledge in the learning process

BackgroundThe identification of robust lists of molecular biomarkers related to a disease is a fundamental step for early diagnosis and treatment. However, methodologies for biomarker discovery using microarray data often provide results with limited overlap. It has been suggested that one reason for these inconsistencies may be that in complex diseases, such as cancer, multiple genes belonging to one or more physiological pathways are associated with the outcomes. Thus, a possible approach to improve list stability is to integrate biological information from genomic databases in the learning process; however, a comprehensive assessment based on different types of biological information is still lacking in the literature. In this work we have compared the effect of using different biological information in the learning process like functional annotations, protein-protein interactions and expression correlation among genes.ResultsBiological knowledge has been codified by means of gene similarity matrices and expression data linearly transformed in such a way that the more similar two features are, the more closely they are mapped. Two semantic similarity matrices, based on Biological Process and Molecular Function Gene Ontology annotation, and geodesic distance applied on protein-protein interaction networks, are the best performers in improving list stability maintaining almost equal prediction accuracy.ConclusionsThe performed analysis supports the idea that when some features are strongly correlated to each other, for example because are close in the protein-protein interaction network, then they might have similar importance and are equally relevant for the task at hand. Obtained results can be a starting point for additional experiments on combining similarity matrices in order to obtain even more stable lists of biomarkers. The implementation of the classification algorithm is available at the link: http://www.math.unipd.it/~dasan/biomarkers.html.

Function-based discovery of significant transcriptional temporal patterns in insulin stimulated muscle cells.

Background Insulin action on protein synthesis (translation of transcripts) and post-translational modifications, especially of those involving the reversible modifications such as phosphorylation of various signaling proteins, are extensively studied but insulin effect on transcription of genes, especially of transcriptional temporal patterns remains to be fully defined. Methodology/Principal Findings To identify significant transcriptional temporal patterns we utilized primary differentiated rat skeletal muscle myotubes which were treated with insulin and samples were collected every 20 min for 8 hours. Pooled samples at every hour were analyzed by gene array approach to measure transcript levels. The patterns of transcript levels were analyzed based on a novel method that integrates selection, clustering, and functional annotation to find the main temporal patterns associated to functional groups of differentially expressed genes. 326 genes were found to be differentially expressed in response to in vitro insulin administration in skeletal muscle myotubes. Approximately 20% of the genes that were differentially expressed were identified as belonging to the insulin signaling pathway. Characteristic transcriptional temporal patterns include: (a) a slow and gradual decrease in gene expression, (b) a gradual increase in gene expression reaching a peak at about 5 hours and then reaching a plateau or an initial decrease and other different variable pattern of increase in gene expression over time. Conclusion/Significance The new method allows identifying characteristic dynamic responses to insulin stimulus, common to a number of genes and associated to the same functional group. The results demonstrate that insulin treatment elicited different clusters of gene transcript profile supporting a temporal regulation of gene expression by insulin in skeletal muscle cells.

Changes in microRNA expression during disease progression in patients with chronic viral hepatitis

MicroRNAs (miRNAs) have been involved in hepatocarcinogenesis, but little is known on their role in the progression of chronic viral hepatitis. Aim of this study was to identify miRNA signatures associated with stages of disease progression in patients with chronic viral hepatitis.

Oncofetal gene SALL4 and prognosis in cancer: A systematic review with meta-analysis

The Spalt-Like Transcription Factor 4 (SALL4) oncogene plays a central function in embryo-fetal development and is absent in differentiated tissues. Evidence suggests that it can be reactivated in several cancers worsening the prognosis. We aimed at investigating the risk associated with SALL4 reactivation for all-cause mortality and recurrence in cancer using the current literature. A PubMed and SCOPUS search until 1st September 2016 was performed, focusing on perspective studies reporting prognostic parameters in cancer data. In addition, 17 datasets of different cancer types from The Cancer Genome Atlas were considered. A total of 9,947 participants across 40 cohorts, followed-up for about 5 years on average, were analyzed comparing patients showing SALL4 presence (SALL4+, n = 1,811) or absence (SALL4-, n = 8,136). All data were summarised using risk ratios (RRs) for the number of deaths/recurrences and hazard ratios (HRs) for the time-dependent risk related to SALL4+, adjusted for potential confounders. SALL4+ significantly increased overall mortality (RR = 1.34, 95% confidence intervals (CI)=1.21-1.48, p<0.0001, I2=66%; HR=1.4; 95%CI: 1.19-1.65; p<0.0001; I2=63%) and recurrence of disease (RR = 1.25, 95% CI = 1.1-1.42, p=0.0006, I2=62%); HR=1.52; 95% CI: 1.22-1.89, p=0.0002; I2=69%) compared to SALL4-. Moreover, SALL4 remained significantly associated with poor prognosis even using HRs adjusted for potential confounders (overall mortality: HR=1.4; 95%CI: 1.19-1.65; p<0.0001; I2=63%; recurrence of disease: HR=1.52; 95% CI: 1.22-1.89, p=0.0002; I2=69%). These results suggest that SALL4 expression increases both mortality and recurrence of cancer, confirming this gene as an important prognostic marker and a potential target for personalized medicine.

The transcriptional response in human umbilical vein endothelial cells exposed to insulin: a dynamic gene expression approach.

Background In diabetes chronic hyperinsulinemia contributes to the instability of the atherosclerotic plaque and stimulates cellular proliferation through the activation of the MAP kinases, which in turn regulate cellular proliferation. However, it is not known whether insulin itself could increase the transcription of specific genes for cellular proliferation in the endothelium. Hence, the characterization of transcriptional modifications in endothelium is an important step for a better understanding of the mechanism of insulin action and the relationship between endothelial cell dysfunction and insulin resistance. Methodology and principal findings The transcriptional response of endothelial cells in the 440 minutes following insulin stimulation was monitored using microarrays and compared to a control condition. About 1700 genes were selected as differentially expressed based on their treated minus control profile, thus allowing the detection of even small but systematic changes in gene expression. Genes were clustered in 7 groups according to their time expression profile and classified into 15 functional categories that can support the biological effects of insulin, based on Gene Ontology enrichment analysis. In terms of endothelial function, the most prominent processes affected were NADH dehydrogenase activity, N-terminal myristoylation domain binding, nitric-oxide synthase regulator activity and growth factor binding. Pathway-based enrichment analysis revealed “Electron Transport Chain” significantly enriched. Results were validated on genes belonging to “Electron Transport Chain” pathway, using quantitative RT-PCR. Conclusions As far as we know, this is the first systematic study in the literature monitoring transcriptional response to insulin in endothelial cells, in a time series microarray experiment. Since chronic hyperinsulinemia contributes to the instability of the atherosclerotic plaque and stimulates cellular proliferation, some of the genes identified in the present work are potential novel candidates in diabetes complications related to endothelial dysfunction.

Knowledge Driven Variable Selection (KDVS) - a new approach to enrichment analysis of gene signatures obtained from high-throughput data

BackgroundHigh–throughput (HT) technologies provide huge amount of gene expression data that can be used to identify biomarkers useful in the clinical practice. The most frequently used approaches first select a set of genes (i.e. gene signature) able to characterize differences between two or more phenotypical conditions, and then provide a functional assessment of the selected genes with an a posteriori enrichment analysis, based on biological knowledge. However, this approach comes with some drawbacks. First, gene selection procedure often requires tunable parameters that affect the outcome, typically producing many false hits. Second, a posteriori enrichment analysis is based on mapping between biological concepts and gene expression measurements, which is hard to compute because of constant changes in biological knowledge and genome analysis. Third, such mapping is typically used in the assessment of the coverage of gene signature by biological concepts, that is either score–based or requires tunable parameters as well, limiting its power.ResultsWe present Knowledge Driven Variable Selection (KDVS), a framework that uses a priori biological knowledge in HT data analysis. The expression data matrix is transformed, according to prior knowledge, into smaller matrices, easier to analyze and to interpret from both computational and biological viewpoints. Therefore KDVS, unlike most approaches, does not exclude a priori any function or process potentially relevant for the biological question under investigation. Differently from the standard approach where gene selection and functional assessment are applied independently, KDVS embeds these two steps into a unified statistical framework, decreasing the variability derived from the threshold–dependent selection, the mapping to the biological concepts, and the signature coverage. We present three case studies to assess the usefulness of the method.ConclusionsWe showed that KDVS not only enables the selection of known biological functionalities with accuracy, but also identification of new ones. An efficient implementation of KDVS was devised to obtain results in a fast and robust way. Computing time is drastically reduced by the effective use of distributed resources. Finally, integrated visualization techniques immediately increase the interpretability of results. Overall, KDVS approach can be considered as a viable alternative to enrichment–based approaches.

MiR-21 over-expression and Programmed Cell Death 4 down-regulation features malignant pleural mesothelioma

Background Differential diagnosis between malignant pleural mesothelioma (MPM) and benign mesothelial conditions is still challenging and there is a lack of useful markers. Programmed cell death 4 (PDCD4) is a well-known tumor suppressor gene in several cancers, its post-transcriptional activity is directly controlled by miR-21, whose over-expression has been recently reported in MPM compared to normal mesothelium. Aim of this study was to test this suppressor gene as a possible new marker of malignant transformation in mesothelial cells, as well as a new prognostic marker. Methods PDCD4 nuclear expression was assessed by immunohistochemistry (IHC) in 40 non-neoplastic pleural (NNP) and 40 MPM formalin-fixed and paraffin-embedded specimens. PDCD4 and miR-21 expressions were analyzed by qRT-PCR in all cases. In situ hybridization (ISH) of miR-21 was performed in 5 representative cases of both groups. The prognostic relevance of PDCD4 was assessed in a public available gene expression dataset. Results IHC showed that PDCD4 nuclear expression was significantly lower in MPM than in NNP. PDCD4 was down-regulated, whereas miR-21 was over-expressed in MPM cases compared to NNP ones. ISH detected miR-21 only in MPM specimens. Down-expression of PDCD4 was found significantly associated with short overall survival in publicly available data. Conclusions These findings highlighted a switch between PDCD4 and miR-21 expression in MPM. Further studies should assess the diagnostic reliability of these two markers for MPM in biopsy and effusion specimens.

DNA methylation profiling reveals common signatures of tumorigenesis and defines epigenetic prognostic subtypes of canine Diffuse Large B-cell Lymphoma

Epigenetic deregulation is a hallmark of cancer characterized by frequent acquisition of new DNA methylation in CpG islands. To gain insight into the methylation changes of canine DLBCL, we investigated the DNA methylome in primary DLBCLs in comparison with control lymph nodes by genome-wide CpG microarray. We identified 1,194 target loci showing different methylation levels in tumors compared with controls. The hypermethylated CpG loci included promoter, 5′-UTRs, upstream and exonic regions. Interestingly, targets of polycomb repressive complex in stem cells were mostly affected suggesting that DLBCL shares a stem cell-like epigenetic pattern. Functional analysis highlighted biological processes strongly related to embryonic development, tissue morphogenesis and cellular differentiation, including HOX, BMP and WNT. In addition, the analysis of epigenetic patterns and genome-wide methylation variability identified cDLBCL subgroups. Some of these epigenetic subtypes showed a concordance with the clinical outcome supporting the hypothesis that the accumulation of aberrant epigenetic changes results in a more aggressive behavior of the tumor. Collectively, our results suggest an important role of DNA methylation in DLBCL where aberrancies in transcription factors were frequently observed, suggesting an involvement during tumorigenesis. These findings warrant further investigation to improve cDLBCL prognostic classification and provide new insights on tumor aggressiveness.