Machine learning for hand pose classification from phasic and tonic EMG signals during bimanual activities in virtual reality.

Myoelectric prostheses have recently shown significant promise for restoring hand function in individuals with upper limb loss or deficiencies, driven by advances in machine learning and increasingly accessible bioelectrical signal acquisition devices. Here, we first introduce and validate a novel experimental paradigm using a virtual reality headset equipped with hand-tracking capabilities to facilitate the recordings of synchronized EMG signals and hand pose estimation. Using both the phasic and tonic EMG components of data acquired through the proposed paradigm, we compare hand gesture classification pipelines based on standard signal processing features, convolutional neural networks, and covariance matrices with Riemannian geometry computed from raw or xDAWN-filtered EMG signals. We demonstrate the performance of the latter for gesture classification using EMG signals. We further hypothesize that introducing physiological knowledge in machine learning models will enhance their performances, leading to better myoelectric prosthesis control. We demonstrate the potential of this approach by using the neurophysiological integration of the "move command" to better separate the phasic and tonic components of the EMG signals, significantly improving the performance of sustained posture recognition. These results pave the way for the development of new cutting-edge machine learning techniques, likely refined by neurophysiology, that will further improve the decoding of real-time natural gestures and, ultimately, the control of myoelectric prostheses.

Improving Infinium MethylationEPIC data processing: re-annotation of enhancers and long noncoding RNA genes and benchmarking of normalization methods.

Illumina Infinium DNA Methylation (5mC) arrays are a popular technology for low-cost, high-throughput, genome-scale measurement of 5mC distribution, especially in cancer and other complex diseases. After the success of its HumanMethylation450 array (450k), Illumina released the MethylationEPIC array (850k) featuring increased coverage of enhancers. Despite the widespread use of 850k, analysis of the corresponding data remains suboptimal: it still relies mostly on Illumina's default annotation, which underestimates enhancerss and long noncoding RNAs. Results: We have thus developed an approach, based on the ENCODE and LNCipedia databases, which greatly improves upon Illumina's default annotation of enhancers and long noncoding transcripts. We compared the re-annotated 850k with both 450k and reduced-representation bisulphite sequencing (RRBS), another high-throughput 5mC profiling technology. We found 850k to cover at least three times as many enhancers and long noncoding RNAs as either 450k or RRBS. We further investigated the reproducibility of the three technologies, applying various normalization methods to the 850k data. Most of these methods reduced variability to a level below that of RRBS data. We then used 850k with our new annotation and normalization to profile 5mC changes in breast cancer biopsies. 850k highlighted aberrant enhancer methylation as the predominant feature, in agreement with previous reports. Our study provides an updated processing approach for 850k data, based on refined probe annotation and normalization, allowing for improved analysis of methylation at enhancers and long noncoding RNA genes. Our findings will help to further advance understanding of the DNA methylome in health and disease.

The role of diversity and ensemble learning in credit card fraud detection.

The number of daily credit card transactions is inexorably growing: the e-commerce market expansion and the recent constraints for the Covid-19 pandemic have significantly increased the use of electronic payments. The ability to precisely detect fraudulent transactions is increasingly important, and machine learning models are now a key component of the detection process. Standard machine learning techniques are widely employed, but inadequate for the evolving nature of customers behavior entailing continuous changes in the underlying data distribution. his problem is often tackled by discarding past knowledge, despite its potential relevance in the case of recurrent concepts. Appropriate exploitation of historical knowledge is necessary: we propose a learning strategy that relies on diversity-based ensemble learning and allows to preserve past concepts and reuse them for a faster adaptation to changes. In our experiments, we adopt several state-of-the-art diversity measures and we perform comparisons with various other learning approaches. We assess the effectiveness of our proposed learning strategy on extracts of two real datasets from two European countries, containing more than 30 M and 50 M transactions, provided by our industrial partner, Worldline, a leading company in the field.

Riemannian classification of single-trial surface EEG and sources during checkerboard and navigational images in humans.

OBJECTIVE: Different visual stimuli are classically used for triggering visual evoked potentials comprising well-defined components linked to the content of the displayed image. These evoked components result from the average of ongoing EEG signals in which additive and oscillatory mechanisms contribute to the component morphology. The evoked related potentials often resulted from a mixed situation (power variation and phase-locking) making basic and clinical interpretations difficult. Besides, the grand average methodology produced artificial constructs that do not reflect individual peculiarities. This motivated new approaches based on single-trial analysis as recently used in the brain-computer interface field. APPROACH: We hypothesize that EEG signals may include specific information about the visual features of the displayed image and that such distinctive traits can be identified by state-of-the-art classification algorithms based on Riemannian geometry. The same classification algorithms are also applied to the dipole sources estimated by sLORETA. MAIN RESULTS AND SIGNIFICANCE: We show that our classification pipeline can effectively discriminate between the display of different visual items (Checkerboard versus 3D navigational image) in single EEG trials throughout multiple subjects. The present methodology reaches a single-trial classification accuracy of about 84% and 93% for inter-subject and intra-subject classification respectively using surface EEG. Interestingly, we note that the classification algorithms trained on sLORETA sources estimation fail to generalize among multiple subjects (63%), which may be due to either the average head model used by sLORETA or the subsequent spatial filtering failing to extract discriminative information, but reach an intra-subject classification accuracy of 82%.

Factor-Based Framework for Multivariate and Multi-step-ahead Forecasting of Large Scale Time Series.

State-of-the-art multivariate forecasting methods are restricted to low dimensional tasks, linear dependencies and short horizons. The technological advances (notably the Big data revolution) are instead shifting the focus to problems characterized by a large number of variables, non-linear dependencies and long forecasting horizons. In the last few years, the majority of the best performing techniques for multivariate forecasting have been based on deep-learning models. However, such models are characterized by high requirements in terms of data availability and computational resources and suffer from a lack of interpretability. To cope with the limitations of these methods, we propose an extension to the DFML framework, a hybrid forecasting technique inspired by the Dynamic Factor Model (DFM) approach, a successful forecasting methodology in econometrics. This extension improves the capabilities of the DFM approach, by implementing and assessing both linear and non-linear factor estimation techniques as well as model-driven and data-driven factor forecasting techniques. We assess several method integrations within the DFML, and we show that the proposed technique provides competitive results both in terms of forecasting accuracy and computational efficiency on multiple very large-scale (>102 variables and > 103 samples) real forecasting tasks.

The CLAIRE COVID-19 initiative: approach, experiences and recommendations.

A volunteer effort by Artificial Intelligence (AI) researchers has shown it can deliver significant research outcomes rapidly to help tackle COVID-19. Within two months, CLAIRE's self-organising volunteers delivered the World's first comprehensive curated repository of COVID-19-related datasets useful for drug-repurposing, drafted review papers on the role CT/X-ray scan analysis and robotics could play, and progressed research in other areas. Given the pace required and nature of voluntary efforts, the teams faced a number of challenges. These offer insights in how better to prepare for future volunteer scientific efforts and large scale, data-dependent AI collaborations in general. We offer seven recommendations on how to best leverage such efforts and collaborations in the context of managing future crises.

Interpreting pathways to discover cancer driver genes with Moonlight.

Cancer driver gene alterations influence cancer development, occurring in oncogenes, tumor suppressors, and dual role genes. Discovering dual role cancer genes is difficult because of their elusive context-dependent behavior. We define oncogenic mediators as genes controlling biological processes. With them, we classify cancer driver genes, unveiling their roles in cancer mechanisms. To this end, we present Moonlight, a tool that incorporates multiple -omics data to identify critical cancer driver genes. With Moonlight, we analyze 8000+ tumor samples from 18 cancer types, discovering 3310 oncogenic mediators, 151 having dual roles. By incorporating additional data (amplification, mutation, DNA methylation, chromatin accessibility), we reveal 1000+ cancer driver genes, corroborating known molecular mechanisms. Additionally, we confirm critical cancer driver genes by analysing cell-line datasets. We discover inactivation of tumor suppressors in intron regions and that tissue type and subtype indicate dual role status. These findings help explain tumor heterogeneity and could guide therapeutic decisions.

Hyperscanning EEG and Classification Based on Riemannian Geometry for Festive and Violent Mental State Discrimination.

Interactions between two brains constitute the essence of social communication. Daily movements are commonly executed during social interactions and are determined by different mental states that may express different positive or negative behavioral intent. In this context, the effective recognition of festive or violent intent before the action execution remains crucial for survival. Here, we hypothesize that the EEG signals contain the distinctive features characterizing movement intent already expressed before movement execution and that such distinctive information can be identified by state-of-the-art classification algorithms based on Riemannian geometry. We demonstrated for the first time that a classifier based on covariance matrices and Riemannian geometry can effectively discriminate between neutral, festive, and violent mental states only on the basis of non-invasive EEG signals in both the actor and observer participants. These results pave the way for new electrophysiological discrimination of mental states based on non-invasive EEG recordings and cutting-edge machine learning techniques.

New functionalities in the TCGAbiolinks package for the study and integration of cancer data from GDC and GTEx.

The advent of Next-Generation Sequencing (NGS) technologies has opened new perspectives in deciphering the genetic mechanisms underlying complex diseases. Nowadays, the amount of genomic data is massive and substantial efforts and new tools are required to unveil the information hidden in the data. The Genomic Data Commons (GDC) Data Portal is a platform that contains different genomic studies including the ones from The Cancer Genome Atlas (TCGA) and the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) initiatives, accounting for more than 40 tumor types originating from nearly 30000 patients. Such platforms, although very attractive, must make sure the stored data are easily accessible and adequately harmonized. Moreover, they have the primary focus on the data storage in a unique place, and they do not provide a comprehensive toolkit for analyses and interpretation of the data. To fulfill this urgent need, comprehensive but easily accessible computational methods for integrative analyses of genomic data that do not renounce a robust statistical and theoretical framework are required. In this context, the R/Bioconductor package TCGAbiolinks was developed, offering a variety of bioinformatics functionalities. Here we introduce new features and enhancements of TCGAbiolinks in terms of i) more accurate and flexible pipelines for differential expression analyses, ii) different methods for tumor purity estimation and filtering, iii) integration of normal samples from other platforms iv) support for other genomics datasets, exemplified here by the TARGET data. Evidence has shown that accounting for tumor purity is essential in the study of tumorigenesis, as these factors promote confounding behavior regarding differential expression analysis. With this in mind, we implemented these filtering procedures in TCGAbiolinks. Moreover, a limitation of some of the TCGA datasets is the unavailability or paucity of corresponding normal samples. We thus integrated into TCGAbiolinks the possibility to use normal samples from the Genotype-Tissue Expression (GTEx) project, which is another large-scale repository cataloging gene expression from healthy individuals. The new functionalities are available in the TCGAbiolinks version 2.8 and higher released in Bioconductor version 3.7.

Comprehensive identification of long noncoding RNAs in colorectal cancer.

Colorectal cancer (CRC) is one of the most common cancers in humans and a leading cause of cancer-related deaths worldwide. As in the case of other cancers, CRC heterogeneity leads to a wide range of clinical outcomes and complicates therapy. Over the years, multiple factors have emerged as markers of CRC heterogeneity, improving tumor classification and selection of therapeutic strategies. Understanding the molecular mechanisms underlying this heterogeneity remains a major challenge. A considerable research effort is therefore devoted to identifying additional features of colorectal tumors, in order to better understand CRC etiology and to multiply therapeutic avenues. Recently, long noncoding RNAs (lncRNAs) have emerged as important players in physiological and pathological processes, including CRC. Here we looked for lncRNAs that might contribute to the various colorectal tumor phenotypes. We thus monitored the expression of 4898 lncRNA genes across 566 CRC samples and identified 282 lncRNAs reflecting CRC heterogeneity. We then inferred potential functions of these lncRNAs. Our results highlight lncRNAs that may participate in the major processes altered in distinct CRC cases, such as WNT/ß-catenin and TGF-ß signaling, immunity, the epithelial-to-mesenchymal transition (EMT), and angiogenesis. For several candidates, we provide experimental evidence supporting our functional predictions that they may be involved in the cell cycle or the EMT. Overall, our work identifies lncRNAs associated with key CRC characteristics and provides insights into their respective functions. Our findings constitute a further step towards understanding the contribution of lncRNAs to CRC heterogeneity. They may open new therapeutic opportunities.

In-Silico Integration Approach to Identify a Key miRNA Regulating a Gene Network in Aggressive Prostate Cancer.

Like other cancer diseases, prostate cancer (PC) is caused by the accumulation of genetic alterations in the cells that drives malignant growth. These alterations are revealed by gene profiling and copy number alteration (CNA) analysis. Moreover, recent evidence suggests that also microRNAs have an important role in PC development. Despite efforts to profile PC, the alterations (gene, CNA, and miRNA) and biological processes that correlate with disease development and progression remain partially elusive. Many gene signatures proposed as diagnostic or prognostic tools in cancer poorly overlap. The identification of co-expressed genes, that are functionally related, can identify a core network of genes associated with PC with a better reproducibility. By combining different approaches, including the integration of mRNA expression profiles, CNAs, and miRNA expression levels, we identified a gene signature of four genes overlapping with other published gene signatures and able to distinguish, in silico, high Gleason-scored PC from normal human tissue, which was further enriched to 19 genes by gene co-expression analysis. From the analysis of miRNAs possibly regulating this network, we found that hsa-miR-153 was highly connected to the genes in the network. Our results identify a four-gene signature with diagnostic and prognostic value in PC and suggest an interesting gene network that could play a key regulatory role in PC development and progression. Furthermore, hsa-miR-153, controlling this network, could be a potential biomarker for theranostics in high Gleason-scored PC.

Integration of multiple networks and pathways identifies cancer driver genes in pan-cancer analysis.

BACKGROUND: Modern high-throughput genomic technologies represent a comprehensive hallmark of molecular changes in pan-cancer studies. Although different cancer gene signatures have been revealed, the mechanism of tumourigenesis has yet to be completely understood. Pathways and networks are important tools to explain the role of genes in functional genomic studies. However, few methods consider the functional non-equal roles of genes in pathways and the complex gene-gene interactions in a network. RESULTS: We present a novel method in pan-cancer analysis that identifies de-regulated genes with a functional role by integrating pathway and network data. A pan-cancer analysis of 7158 tumour/normal samples from 16 cancer types identified 895 genes with a central role in pathways and de-regulated in cancer. Comparing our approach with 15 current tools that identify cancer driver genes, we found that 35.6% of the 895 genes identified by our method have been found as cancer driver genes with at least 2/15 tools. Finally, we applied a machine learning algorithm on 16 independent GEO cancer datasets to validate the diagnostic role of cancer driver genes for each cancer. We obtained a list of the top-ten cancer driver genes for each cancer considered in this study. CONCLUSIONS: Our analysis 1) confirmed that there are several known cancer driver genes in common among different types of cancer, 2) highlighted that cancer driver genes are able to regulate crucial pathways.

Credit Card Fraud Detection: A Realistic Modeling and a Novel Learning Strategy.

Detecting frauds in credit card transactions is perhaps one of the best testbeds for computational intelligence algorithms. In fact, this problem involves a number of relevant challenges, namely: concept drift (customers' habits evolve and fraudsters change their strategies over time), class imbalance (genuine transactions far outnumber frauds), and verification latency (only a small set of transactions are timely checked by investigators). However, the vast majority of learning algorithms that have been proposed for fraud detection rely on assumptions that hardly hold in a real-world fraud-detection system (FDS). This lack of realism concerns two main aspects: 1) the way and timing with which supervised information is provided and 2) the measures used to assess fraud-detection performance. This paper has three major contributions. First, we propose, with the help of our industrial partner, a formalization of the fraud-detection problem that realistically describes the operating conditions of FDSs that everyday analyze massive streams of credit card transactions. We also illustrate the most appropriate performance measures to be used for fraud-detection purposes. Second, we design and assess a novel learning strategy that effectively addresses class imbalance, concept drift, and verification latency. Third, in our experiments, we demonstrate the impact of class unbalance and concept drift in a real-world data stream containing more than 75 million transactions, authorized over a time window of three years.

Combination of Gene Expression Signature and Model for End-Stage Liver Disease Score Predicts Survival of Patients With Severe Alcoholic Hepatitis.

BACKGROUND & AIMS: Patients with severe alcoholic hepatitis (AH) have a high risk of death within 90 days. Corticosteroids, which can cause severe adverse events, are the only treatment that increases short-term survival. It is a challenge to predict outcomes of patients with severe AH. Therefore, we developed a scoring system to predict patient survival, integrating baseline molecular and clinical variables. METHODS: We obtained fixed liver biopsy samples from 71 consecutive patients diagnosed with severe AH and treated with corticosteroids from July 2006 through December 2013 in Brussels, Belgium (derivation cohort). Gene expression patterns were analyzed by microarrays and clinical data were collected for 180 days. We identified gene expression signatures and clinical data that are associated with survival without liver transplantation at 90 and 180 days after initiation of corticosteroid therapy. Findings were validated using liver biopsies from 48 consecutive patients with severe AH treated with corticosteroids, collected from March 2010 through February 2015 at hospitals in Belgium and Switzerland (validation cohort 1) and in liver biopsies from 20 patients (9 received corticosteroid treatment), collected from January 2012 through May 2015 in the United States (validation cohort 2). RESULTS: We integrated data on expression patterns of 123 genes and the model for end-stage liver disease (MELD) scores to assign patients to groups with poor survival (29% survived 90 days and 26% survived 180 days) and good survival (76% survived 90 days and 65% survived 180 days) (P < .001) in the derivation cohort. We named this assignment system the gene signature-MELD (gs-MELD) score. In validation cohort 1, the gs-MELD score discriminated patients with poor survival (43% survived 90 days) from those with good survival (96% survived 90 days) (P < .001). The gs-MELD score also discriminated between patients with a poor survival at 180 days (34% survived) and a good survival at 180 days (84% survived) (P < .001). The time-dependent area under the receiver operator characteristic curve for the score was 0.86 (95% confidence interval 0.73-0.99) for survival at 90 days, and 0.83 (95% confidence interval 0.71-0.96) for survival at 180 days. This score outperformed other clinical models to predict survival of patients with severe AH in validation cohort 1. In validation cohort 2, the gs-MELD discriminated patients with a poor survival at 90 days (12% survived) from those with a good survival at 90 days (100%) (P < .001). CONCLUSIONS: We integrated data on baseline liver gene expression pattern and the MELD score to create the gs-MELD scoring system, which identifies patients with severe AH, treated or not with corticosteroids, most and least likely to survive for 90 and 180 days.

DNA methylation-based immune response signature improves patient diagnosis in multiple cancers.

BACKGROUND: The tumor immune response is increasingly associated with better clinical outcomes in breast and other cancers. However, the evaluation of tumor-infiltrating lymphocytes (TILs) relies on histopathological measurements with limited accuracy and reproducibility. Here, we profiled DNA methylation markers to identify a methylation of TIL (MeTIL) signature that recapitulates TIL evaluations and their prognostic value for long-term outcomes in breast cancer (BC). METHODS: MeTIL signature scores were correlated with clinical endpoints reflecting overall or disease-free survival and a pathologic complete response to preoperative anthracycline therapy in 3 BC cohorts from the Jules Bordet Institute in Brussels and in other cancer types from The Cancer Genome Atlas. RESULTS: The MeTIL signature measured TIL distributions in a sensitive manner and predicted survival and response to chemotherapy in BC better than did histopathological assessment of TILs or gene expression-based immune markers, respectively. The MeTIL signature also improved the prediction of survival in other malignancies, including melanoma and lung cancer. Furthermore, the MeTIL signature predicted differences in survival for malignancies in which TILs were not known to have a prognostic value. Finally, we showed that MeTIL markers can be determined by bisulfite pyrosequencing of small amounts of DNA from formalin-fixed, paraffin-embedded tumor tissue, supporting clinical applications for this methodology. CONCLUSIONS: This study highlights the power of DNA methylation to evaluate tumor immune responses and the potential of this approach to improve the diagnosis and treatment of breast and other cancers. FUNDING: This work was funded by the Fonds National de la Recherche Scientifique (FNRS) and Télévie, the INNOVIRIS Brussels Region BRUBREAST Project, the IUAP P7/03 program, the Belgian "Foundation against Cancer," the Breast Cancer Research Foundation (BCRF), and the Fonds Gaston Ithier.

Novel promoters and coding first exons in DLG2 linked to developmental disorders and intellectual disability.

BACKGROUND: Tissue-specific integrative omics has the potential to reveal new genic elements important for developmental disorders. METHODS: Two pediatric patients with global developmental delay and intellectual disability phenotype underwent array-CGH genetic testing, both showing a partial deletion of the DLG2 gene. From independent human and murine omics datasets, we combined copy number variations, histone modifications, developmental tissue-specific regulation, and protein data to explore the molecular mechanism at play. RESULTS: Integrating genomics, transcriptomics, and epigenomics data, we describe two novel DLG2 promoters and coding first exons expressed in human fetal brain. Their murine conservation and protein-level evidence allowed us to produce new DLG2 gene models for human and mouse. These new genic elements are deleted in 90% of 29 patients (public and in-house) showing partial deletion of the DLG2 gene. The patients' clinical characteristics expand the neurodevelopmental phenotypic spectrum linked to DLG2 gene disruption to cognitive and behavioral categories. CONCLUSIONS: While protein-coding genes are regarded as well known, our work shows that integration of multiple omics datasets can unveil novel coding elements. From a clinical perspective, our work demonstrates that two new DLG2 promoters and exons are crucial for the neurodevelopmental phenotypes associated with this gene. In addition, our work brings evidence for the lack of cross-annotation in human versus mouse reference genomes and nucleotide versus protein databases.

CancerSubtypes: an R/Bioconductor package for molecular cancer subtype identification, validation and visualization.

SUMMARY: Identifying molecular cancer subtypes from multi-omics data is an important step in the personalized medicine. We introduce CancerSubtypes, an R package for identifying cancer subtypes using multi-omics data, including gene expression, miRNA expression and DNA methylation data. CancerSubtypes integrates four main computational methods which are highly cited for cancer subtype identification and provides a standardized framework for data pre-processing, feature selection, and result follow-up analyses, including results computing, biology validation and visualization. The input and output of each step in the framework are packaged in the same data format, making it convenience to compare different methods. The package is useful for inferring cancer subtypes from an input genomic dataset, comparing the predictions from different well-known methods and testing new subtype discovery methods, as shown with different application scenarios in the Supplementary Material. AVAILABILITY AND IMPLEMENTATION: The package is implemented in R and available under GPL-2 license from the Bioconductor website (http://bioconductor.org/packages/CancerSubtypes/). CONTACT: thuc.le@unisa.edu.au or jiuyong.li@unisa.edu.au. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Study of Meta-analysis strategies for network inference using information-theoretic approaches.

BACKGROUND: Reverse engineering of gene regulatory networks (GRNs) from gene expression data is a classical challenge in systems biology. Thanks to high-throughput technologies, a massive amount of gene-expression data has been accumulated in the public repositories. Modelling GRNs from multiple experiments (also called integrative analysis) has; therefore, naturally become a standard procedure in modern computational biology. Indeed, such analysis is usually more robust than the traditional approaches, which suffer from experimental biases and the low number of samples by analysing individual datasets. To date, there are mainly two strategies for the problem of interest: the first one ("data merging") merges all datasets together and then infers a GRN whereas the other ("networks ensemble") infers GRNs from every dataset separately and then aggregates them using some ensemble rules (such as ranksum or weightsum). Unfortunately, a thorough comparison of these two approaches is lacking. RESULTS: In this work, we are going to present another meta-analysis approach for inferring GRNs from multiple studies. Our proposed meta-analysis approach, adapted to methods based on pairwise measures such as correlation or mutual information, consists of two steps: aggregating matrices of the pairwise measures from every dataset followed by extracting the network from the meta-matrix. Afterwards, we evaluate the performance of the two commonly used approaches mentioned above and our presented approach with a systematic set of experiments based on in silico benchmarks. CONCLUSIONS: We proposed a first systematic evaluation of different strategies for reverse engineering GRNs from multiple datasets. Experiment results strongly suggest that assembling matrices of pairwise dependencies is a better strategy for network inference than the two commonly used ones.

SpidermiR: An R/Bioconductor Package for Integrative Analysis with miRNA Data.

Gene Regulatory Networks (GRNs) control many biological systems, but how such network coordination is shaped is still unknown. GRNs can be subdivided into basic connections that describe how the network members interact e.g., co-expression, physical interaction, co-localization, genetic influence, pathways, and shared protein domains. The important regulatory mechanisms of these networks involve miRNAs. We developed an R/Bioconductor package, namely SpidermiR, which offers an easy access to both GRNs and miRNAs to the end user, and integrates this information with differentially expressed genes obtained from The Cancer Genome Atlas. Specifically, SpidermiR allows the users to: (i) query and download GRNs and miRNAs from validated and predicted repositories; (ii) integrate miRNAs with GRNs in order to obtain miRNA-gene-gene and miRNA-protein-protein interactions, and to analyze miRNA GRNs in order to identify miRNA-gene communities; and (iii) graphically visualize the results of the analyses. These analyses can be performed through a single interface and without the need for any downloads. The full data sets are then rapidly integrated and processed locally.

Portraying breast cancers with long noncoding RNAs.

Evidence is emerging that long noncoding RNAs (lncRNAs) may play a role in cancer development, but this role is not yet clear. We performed a genome-wide transcriptional survey to explore the lncRNA landscape across 995 breast tissue samples. We identified 215 lncRNAs whose genes are aberrantly expressed in breast tumors, as compared to normal samples. Unsupervised hierarchical clustering of breast tumors on the basis of their lncRNAs revealed four breast cancer subgroups that correlate tightly with PAM50-defined mRNA-based subtypes. Using multivariate analysis, we identified no less than 210 lncRNAs prognostic of clinical outcome. By analyzing the coexpression of lncRNA genes and protein-coding genes, we inferred potential functions of the 215 dysregulated lncRNAs. We then associated subtype-specific lncRNAs with key molecular processes involved in cancer. A correlation was observed, on the one hand, between luminal A-specific lncRNAs and the activation of phosphatidylinositol 3-kinase, fibroblast growth factor, and transforming growth factor-ß pathways and, on the other hand, between basal-like-specific lncRNAs and the activation of epidermal growth factor receptor (EGFR)-dependent pathways and of the epithelial-to-mesenchymal transition. Finally, we showed that a specific lncRNA, which we called CYTOR, plays a role in breast cancer. We confirmed its predicted functions, showing that it regulates genes involved in the EGFR/mammalian target of rapamycin pathway and is required for cell proliferation, cell migration, and cytoskeleton organization. Overall, our work provides the most comprehensive analyses for lncRNA in breast cancers. Our findings suggest a wide range of biological functions associated with lncRNAs in breast cancer and provide a foundation for functional investigations that could lead to new therapeutic approaches.