Search

Gene expression data holds the potential to shed light on multiple biological processes at once. However, data analysis methods for single cell sequencing mostly focus on finding cell clusters or the principal progression line of the data. Data analysis for spatial transcriptomics mostly addresses clustering and finding spatially variable genes. Existing data analysis methods are effective in finding the main data features, but they might miss less pronounced, albeit significant, processes, possibly involving a subset of the samples. In this work we present SPIRAL: Significant Process InfeRence ALgorithm. SPIRAL is based on Gaussian statistics to detect all statistically significant biological processes in single cell, bulk and spatial transcriptomics data. The algorithm outputs a list of structures, each defined by a set of genes working simultaneously in a specific population of cells. SPIRAL is unique in its flexibility: the structures are constructed by selecting subsets of genes and cells based on statistically significant and consistent differential expression. Every gene and every cell may be part of one structure, more or none. SPIRAL also provides several visual representations of structures and pathway enrichment information. We validated the statistical soundness of SPIRAL on synthetic datasets and applied it to single cell, spatial and bulk RNA-sequencing datasets. SPIRAL is available at https://spiral.technion.ac.il/ ., (© 2024. The Author(s).)

Different miRNA profiling protocols and technologies introduce differences in the resulting quantitative expression profiles. These include differences in the presence (and measurability) of certain miRNAs. We present and examine a method based on quantile normalization, Adjusted Quantile Normalization (AQuN), to combine miRNA expression data from multiple studies in breast cancer into a single joint dataset for integrative analysis. By pooling multiple datasets, we obtain increased statistical power, surfacing patterns that do not emerge as statistically significant when separately analyzing these datasets. To merge several datasets, as we do here, one needs to overcome both technical and batch differences between these datasets. We compare several approaches for merging and jointly analyzing miRNA datasets. We investigate the statistical confidence for known results and highlight potential new findings that resulted from the joint analysis using AQuN. In particular, we detect several miRNAs to be differentially expressed in estrogen receptor (ER) positive versus ER negative samples. In addition, we identify new potential biomarkers and therapeutic targets for both clinical groups. As a specific example, using the AQuN-derived dataset we detect hsa-miR-193b-5p to have a statistically significant over-expression in the ER positive group, a phenomenon that was not previously reported. Furthermore, as demonstrated by functional assays in breast cancer cell lines, overexpression of hsa-miR-193b-5p in breast cancer cell lines resulted in decreased cell viability in addition to inducing apoptosis. Together, these observations suggest a novel functional role for this miRNA in breast cancer. Packages implementing AQuN are provided for Python and Matlab: https://github.com/YakhiniGroup/PyAQN., Competing Interests: The authors have declared that no competing interests exist.

ENViz (Enrichment Analysis and Visualization) is a Cytoscape app that performs joint enrichment analysis of two types of sample matched datasets in the context of systematic annotations. Such datasets may be gene expression or any other high-throughput data collected in the same set of samples. The enrichment analysis is done in the context of pathway information, gene ontology or any custom annotation of the data. The results of the analysis consist of significant associations between profiled elements of one of the datasets to the annotation terms (e.g. miR-19 was associated to the cell-cycle process in breast cancer samples). The results of the enrichment analysis are visualized as an interactive Cytoscape network., (© The Author 2015. Published by Oxford University Press.)

Time to freezing tumor tissue for RNA expression analysis will always vary to some extent. To evaluate the effect of ischemia time, tumor tissue from ten breast cancer patients was collected and aliquots of tissue were snap frozen at different time points after surgery (0, 0.5, 1, 3 and 6 h). Using miRNA and mRNA expression microarrays and statistical analysis, 56 miRNAs and 1788 mRNAs were found to be significantly altered with ischemia time up to six hours. Several of the 56 miRNAs have been reported to play a role in cancer, such as hsa-miR-663 and hsa-miR-125a-3p. Known stress response genes such as GADD45B, JUND and FOSB were among the mRNAs most significantly affected by time to freezing. A novel statistical method for identification of consistently correlated miRNA-mRNA pairs and miRNA-associated biological processes in time course data is presented. Application of this method revealed that several miRNAs, including hsa-miR-1228, hsa-miR-1225-5p and hsa-miR-574-5p, were associated through their correlation to mRNAs to biological processes such as "response to stimulus" and "stress response". These miRNAs also showed enrichment of predicted targets among either their positively or negatively correlated mRNAs. The induced miRNAs may play both direct and indirect roles in biological responses. Caution should be taken when the miRNAs and mRNAs reported to be affected by ischemia time are included in a prognostic or predictive signature., (Copyright © 2011 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.)

Background: MicroRNAs (miRNAs) regulate target genes at the post-transcriptional level and play important roles in cancer pathogenesis and development. Variation amongst individuals is a significant confounding factor in miRNA (or other) expression studies. The true character of biologically or clinically meaningful differential expression can be obscured by inter-patient variation. In this study we aim to identify miRNAs with consistent differential expression in multiple tumor types using a novel data analysis approach., Methods: Using microarrays we profiled the expression of more than 700 miRNAs in 28 matched tumor/normal samples from 8 different tumor types (breast, colon, liver, lung, lymphoma, ovary, prostate and testis). This set is unique in putting emphasis on minimizing tissue type and patient related variability using normal and tumor samples from the same patient. We develop scores for comparing miRNA expression in the above matched sample data based on a rigorous characterization of the distribution of order statistics over a discrete state set, including exact p-values. Specifically, we compute a Rank Consistency Score (RCoS) for every miRNA measured in our data. Our methods are also applicable in various other contexts. We compare our methods, as applied to matched samples, to paired t-test and to the Wilcoxon Signed Rank test., Results: We identify consistent (across the cancer types measured) differentially expressed miRNAs. 41 miRNAs are under-expressed in cancer compared to normal, at FDR (False Discovery Rate) of 0.05 and 17 are over-expressed at the same FDR level. Differentially expressed miRNAs include known oncomiRs (e.g miR-96) as well as miRNAs that were not previously universally associated with cancer. Specific examples include miR-133b and miR-486-5p, which are consistently down regulated and mir-629* which is consistently up regulated in cancer, in the context of our cohort. Data is available in GEO. Software is available at: http://bioinfo.cs.technion.ac.il/people/zohar/RCoS/

Motivation: Unsupervised class discovery in gene expression data relies on the statistical signals in the data to exclusively drive the results. It is often the case, however, that one is interested in constraining the search space to respect certain biological prior knowledge while still allowing a flexible search within these boundaries., Results: We develop an approach to semi-supervised class discovery. One component of our approach uses clinical sample information to constrain the search space and guide the class discovery process to yield biologically relevant partitions. A second component consists of using known biological annotation of genes to drive the search, seeking partitions that manifest strong differential expression in specific sets of genes. We develop efficient algorithmics for these tasks, implementing both approaches and combinations thereof. We show that our method is robust enough to detect known clinical parameters in accordance with expected clinical values. We also use our method to elucidate cardiovascular disease (CVD) putative risk factors., Availability: MonoClaD (Monotone Class Discovery). See http:// bioinfo.cs.technion.ac.il/people/zohar/MonoClad/., Supplementary Information: Supplementary data is available at http://bioinfo.cs.technion.ac.il/people/zohar/MonoClad/software. html

Motivation: Large-scale association studies, investigating the genetic determinants of a phenotype of interest, are producing increasing amounts of genomic variation data on human cohorts. A fundamental challenge in these studies is the detection of genotypic patterns that discriminate individuals exhibiting the phenotype under study from individuals that do not possess it. The difficulty stems from the large number of single nucleotide polymorphism (SNP) combinations that have to be tested. The discrimination problem becomes even more involved when additional high-throughput data, such as gene expression data, are available for the same cohort., Results: We have developed a graph theoretic approach for identifying discriminating patterns (DPs) for a given phenotype in a genotyped population. The method is based on representing the SNP data as a bipartite graph of individuals and their SNP states, and identifying fully connected subgraphs of this graph that relate individuals enriched for a given phenotypic group. The method can handle additional data types such as expression profiles of the genotyped population. It is reminiscent of biclustering approaches with the crucial difference that its search process is guided by the phenotype under consideration in a supervised manner. We tested our approach in simulations and on real data. In simulations, our method was able to retrieve planted patterns with high success rate. We then applied our approach to a dataset of 72 breast cancer patients with available gene expression profiles, genotyped over 695 SNPs. We detected several DPs that were highly significant with respect to various clinical phenotypes, and investigated the groups of patients and the groups of genes they defined. We found the patient groups to be highly enriched for other phenotypes and to display expression coherency among their profiles. The gene groups displayed functional coherency and involved genes with known role in cancer, providing additional support to their involvement., Availability: The program is available upon request.

Atherosclerosis occurs predominantly in arteries and only rarely in veins. The goal of this study was to test whether differences in the molecular responses of venous and arterial endothelial cells (ECs) to atherosclerotic stimuli might contribute to vascular bed differences in susceptibility to atherosclerosis. We compared gene expression profiles of primary cultured ECs from human saphenous vein (SVEC) and coronary artery (CAEC) exposed to atherogenic stimuli. In addition to identifying differentially expressed genes, we applied statistical analysis of gene ontology and pathway annotation terms to identify signaling differences related to cell type and stimulus. Differential gene expression of untreated venous and arterial endothelial cells yielded 285 genes more highly expressed in untreated SVEC (P<0.005 and fold change >1.5). These genes represented various atherosclerosis-related pathways including responses to proliferation, oxidoreductase activity, antiinflammatory responses, cell growth, and hemostasis functions. Moreover, stimulation with oxidized LDL induced dramatically greater gene expression responses in CAEC compared with SVEC, relating to adhesion, proliferation, and apoptosis pathways. In contrast, interleukin 1beta and tumor necrosis factor alpha activated similar gene expression responses in both CAEC and SVEC. The differences in functional response and gene expression were further validated by an in vitro proliferation assay and in vivo immunostaining of alphabeta-crystallin protein. Our results strongly suggest that different inherent gene expression programs in arterial versus venous endothelial cells contribute to differences in atherosclerotic disease susceptibility.

High throughput expression profiling and genotyping technologies provide the means to study the genetic determinants of population variation in gene expression variation. In this paper we present a general statistical framework for the simultaneous analysis of gene expression data and SNP genotype data measured for the same cohort. The framework consists of methods to associate transcripts with SNPs affecting their expression, algorithms to detect subsets of transcripts that share significantly many associations with a subset of SNPs, and methods to visualize the identified relations. We apply our framework to SNP-expression data collected from 49 breast cancer patients. Our results demonstrate an overabundance of transcript-SNP associations in this data, and pinpoint SNPs that are potential master regulators of transcription. We also identify several statistically significant transcript-subsets with common putative regulators that fall into well-defined functional categories.

Background: Apelin is among the most potent stimulators of cardiac contractility known. However, no physiological or pathological role for apelin-angiotensin receptor-like 1 (APJ) signaling has ever been described., Methods and Results: We performed transcriptional profiling using a spotted cDNA microarray with 12 814 unique clones on paired samples of left ventricle obtained before and after placement of a left ventricular assist device in 11 patients. The significance analysis of microarrays and a novel rank consistency score designed to exploit the paired structure of the data confirmed that natriuretic peptides were among the most significantly downregulated genes after offloading. The most significantly upregulated gene was the G-protein-coupled receptor APJ, the specific receptor for apelin. We demonstrate here using immunoassay and immunohistochemical techniques that apelin is localized primarily in the endothelium of the coronary arteries and is found at a higher concentration in cardiac tissue after mechanical offloading. These findings imply an important paracrine signaling pathway in the heart. We additionally extend the clinical significance of this work by reporting for the first time circulating human apelin levels and demonstrating increases in the plasma level of apelin in patients with left ventricular dysfunction., Conclusions: The apelin-APJ signaling pathway emerges as an important novel mediator of cardiovascular control.

This paper concerns the discovery of patterns in gene expression matrices, in which each element gives the expression level of a given gene in a given experiment. Most existing methods for pattern discovery in such matrices are based on clustering genes by comparing their expression levels in all experiments, or clustering experiments by comparing their expression levels for all genes. Our work goes beyond such global approaches by looking for local patterns that manifest themselves when we focus simultaneously on a subset G of the genes and a subset T of the experiments. Specifically, we look for order-preserving submatrices (OPSMs), in which the expression levels of all genes induce the same linear ordering of the experiments (we show that the OPSM search problem is NP-hard in the worst case). Such a pattern might arise, for example, if the experiments in T represent distinct stages in the progress of a disease or in a cellular process and the expression levels of all genes in G vary across the stages in the same way. We define a probabilistic model in which an OPSM is hidden within an otherwise random matrix. Guided by this model, we develop an efficient algorithm for finding the hidden OPSM in the random matrix. In data generated according to the model, the algorithm recovers the hidden OPSM with a very high success rate. Application of the methods to breast cancer data seem to reveal significant local patterns.

Studies of the expression patterns of many genes simultaneously lead to the observation that even in closely related pathologies, there are numerous genes that are differentially expressed in consistent patterns correlated to each sample type. The early uses of the enabling technology, microarrays, was focused on gathering mechanistic biological insights. The early findings now pose another clear challenge, finding ways to effectively use this kind of information to develop diagnostics.

The most common human cancers are malignant neoplasms of the skin. Incidence of cutaneous melanoma is rising especially steeply, with minimal progress in non-surgical treatment of advanced disease. Despite significant effort to identify independent predictors of melanoma outcome, no accepted histopathological, molecular or immunohistochemical marker defines subsets of this neoplasm. Accordingly, though melanoma is thought to present with different 'taxonomic' forms, these are considered part of a continuous spectrum rather than discrete entities. Here we report the discovery of a subset of melanomas identified by mathematical analysis of gene expression in a series of samples. Remarkably, many genes underlying the classification of this subset are differentially regulated in invasive melanomas that form primitive tubular networks in vitro, a feature of some highly aggressive metastatic melanomas. Global transcript analysis can identify unrecognized subtypes of cutaneous melanoma and predict experimentally verifiable phenotypic characteristics that may be of importance to disease progression.

Constantly improving gene expression profiling technologies are expected to provide understanding and insight into cancer-related cellular processes. Gene expression data is also expected to significantly aid in the development of efficient cancer diagnosis and classification platforms. In this work we examine three sets of gene expression data measured across sets of tumor(s) and normal clinical samples: The first set consists of 2,000 genes, measured in 62 epithelial colon samples (Alon et al., 1999). The second consists of approximately equal to 100,000 clones, measured in 32 ovarian samples (unpublished extension of data set described in Schummer et al. (1999)). The third set consists of approximately equal to 7,100 genes, measured in 72 bone marrow and peripheral blood samples (Golub et al, 1999). We examine the use of scoring methods, measuring separation of tissue type (e.g., tumors from normals) using individual gene expression levels. These are then coupled with high-dimensional classification methods to assess the classification power of complete expression profiles. We present results of performing leave-one-out cross validation (LOOCV) experiments on the three data sets, employing nearest neighbor classifier, SVM (Cortes and Vapnik, 1995), AdaBoost (Freund and Schapire, 1997) and a novel clustering-based classification technique. As tumor samples can differ from normal samples in their cell-type composition, we also perform LOOCV experiments using appropriately modified sets of genes, attempting to eliminate the resulting bias. We demonstrate success rate of at least 90% in tumor versus normal classification, using sets of selected genes, with, as well as without, cellular-contamination-related members. These results are insensitive to the exact selection mechanism, over a certain range.