The potential of applying large-scale multi-omics data to unsupervised learning-based transdiagnostic clusters of individuals with severe mental disorders to identify pharmacodynamic targets.

BackgroundOne major goal of large-scale cancer omics study is to identify molecular subtypes for more accurate cancer diagnoses and treatments. To deal with high-dimensional cancer multi-omics data, a promising strategy is to find an effective low-dimensional subspace of the original data and then cluster cancer samples in the reduced subspace. However, due to data-type diversity and big data volume, few methods can integrative and efficiently find the principal low-dimensional manifold of the high-dimensional cancer multi-omics data.ResultsIn this study, we proposed a novel low-rank approximation based integrative probabilistic model to fast find the shared principal subspace across multiple data types: the convexity of the low-rank regularized likelihood function of the probabilistic model ensures efficient and stable model fitting. Candidate molecular subtypes can be identified by unsupervised clustering hundreds of cancer samples in the reduced low-dimensional subspace. On testing datasets, our method LRAcluster (low-rank approximation based multi-omics data clustering) runs much faster with better clustering performances than the existing method. Then, we applied LRAcluster on large-scale cancer multi-omics data from TCGA. The pan-cancer analysis results show that the cancers of different tissue origins are generally grouped as independent clusters, except squamous-like carcinomas. While the single cancer type analysis suggests that the omics data have different subtyping abilities for different cancer types.ConclusionsLRAcluster is a very useful method for fast dimension reduction and unsupervised clustering of large-scale multi-omics data. LRAcluster is implemented in R and freely available via http://bioinfo.au.tsinghua.edu.cn/software/lracluster/.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-015-2223-8) contains supplementary material, which is available to authorized users.

The Internet-of-Things (IoT) is increasingly becoming a major challenge for network administrators to monitor and manage connected devices and sensors, ranging from smart-lights to smoke-alarms and security-cameras. In addition to new device offerings, manufacturers tend to automatically perform firmware upgrade from their cloud servers to change functionalities of existing devices that are operational in the field. This makes it difficult to re-train device classification models in order to capture legitimate changes dynamically. In this paper, we develop a modular device classification architecture that allows us to dynamically accommodate legitimate changes in network IoT assets, either addition of a new device type or upgrades of existing types, without replacing the entire set of models. Our contributions are twofold: (1) We identify key traffic attributes that can be obtained from flow-level network telemetry to characterize individual IoT devices. We develop an unsupervised one-class clustering method for each device to detect its normal network behavior. (2) We tune individual device-specific clustering models and use them to classify IoT devices in real-time. We enhance our classification by developing methods for automatic conflict resolution and noise filtering. We evaluate the efficacy of our scheme by applying it to traffic traces of ten real IoT devices, and demonstrate its ability to achieve overall accuracy of more than 94%.

Deep learning has been well-known for a couple of years, and it indicates incredible possibilities for unsupervised learning of representations with the clustering algorithm. The forms of Convolution Neural Networks (CNN) are now state-of-the-art for many recognition and clustering tasks. However, with the perpetual incrementation of digital images, there exist more and more redundant, irrelevant, and noisy samples which cause CNN running to gradually decrease, and its clustering accuracy decreases concurrently. To conquer these issues, we proposed an effective clustering method for a large-scale image dataset which combines CNN and a Fuzzy-Rough C-Mean (FRCM) clustering algorithm. The main idea is that first a high-level representation, learned by multi-layers of CNN with one clustering layer, produce the initial cluster center, then during training image clusters, and representations, are updating jointly. FRCM is utilized to update the cluster centers in the forward pass, while the parameters of proposed CNN are updated by the backward pass based on Stochastic Gradient Descent (SGD). The concept of the rough set of lower and boundary approximations deal with uncertainty, vagueness, and incompleteness in cluster definition, and fuzzy sets enable efficient handling of overlapping partitions in the noisy environment. The experiment results show that the proposed FRCM based unsupervised CNN clustering method is better than the standard K-Mean, Fuzzy C-Mean, FRCM and also other deep-learning-based clustering algorithms on large-scale image data.

BackgroundSepsis is a major global health problem. However, it lacks a true reference standard for case identification, complicating epidemiologic surveillance. Consensus definitions have changed multiple times, clinicians struggle to identify sepsis at the bedside, and differing identification algorithms generate wide variation in incidence rates. The two current identification approaches use codes from administrative data, or electronic health record (EHR)-based algorithms such as the Center for Disease Control Adult Sepsis Event (ASE); both have limitations. Here our primary purpose is to report initial steps in developing a novel approach to identifying sepsis using unsupervised clustering methods. Secondarily, we report preliminary analysis of resulting clusters, using identification by ASE criteria as a familiar comparator.MethodsThis retrospective cohort study used hospital administrative and EHR data on adults admitted to intensive care units (ICUs) at five Canadian medical centres (2015–2017), with split development and validation cohorts. After preprocessing 592 variables (demographics, encounter characteristics, diagnoses, medications, laboratory tests, and clinical management) and applying data reduction, we presented 55 principal components to eight different clustering algorithms. An automated elbow method determined the optimal number of clusters, and the optimal algorithm was selected based on clustering metrics for consistency, separation, distribution and stability. Cluster membership in the validation cohort was assigned using an XGBoost model trained to predict cluster membership in the development cohort. For cluster analysis, we prospectively subdivided clusters by their fractions meeting ASE criteria (≥ 50% ASE-majority clusters vs. ASE-minority clusters), and compared their characteristics.ResultsThere were 3660 patients in the development cohort and 3012 in the validation cohort, of which 21.5% (development) and 19.1% (validation) were ASE (+). The Robust and Sparse K-means Clustering (RSKC) method performed best. In the development cohort, it identified 48 clusters of hospitalizations; 11 ASE-majority clusters contained 22.4% of all patients but 77.8% of all ASE (+) patients. 34.9% of the 209 ASE (−) patients in the ASE-majority clusters met more liberal ASE criteria for sepsis. Findings were consistent in the validation cohort.ConclusionsUnsupervised clustering applied to diverse, large-scale medical data offers a promising approach to the identification of sepsis phenotypes for epidemiological surveillance.

Social Networks have been developed dramatically just in decades. People now have a convenient way to interact with both social media and other people by making the most of using these social networks. Nevertheless, there is still lack of useful tools that can help users (both consumers and providers) managing such social media under events perspective. In order to meet one of these emerging requirements, a user-centric parallel split-n-merge framework applied for un-supervised clustering social media events is introduced. The purpose of this framework is to cluster social media to events they depict by exploiting and exploring the role of users (who) and the way users interact with data (where, what, when) and others (what, who). The output of the proposed framework can be used for event organization/summarization, and as pre-processing stage for event detection and tracking. Major advantages of the proposed framework are (1) low computational solution w.r.t large-scale data, (2) parallel running, and (3) unsupervised clustering with no training data and third-party information requirements. The comparison between the proposed framework and up-to-date methods with MediaEval2013 1 test-bed and evaluation tools shows a very competitive result.

Introduction Breast cancer is the most frequently diagnosed malignancy in women worldwide. Heterogeneity is a characteristic of tumour aggression and metastatic progression in breast cancer. With the power of single-cell analysis we uncovered key subpopulations within this heterogeneous landscape and specific progressive metastatic characteristics. Methods Single cell RNA sequencing was carried on three endocrine resistant ER positive xenograft tumours with varying levels of metastatic burden. Data analysis was performed using the Seurat pipeline for filtering, horizontal data integration, cell cycle regression and unsupervised clustering. Individual clusters were characterized through differentially expressed markers and PAM50 molecular subtypes distribution. Results Data integration of scRNA sequencing from primary tumours revealed seven unsupervised clusters, six of which are present in all three mice, highlighting the strong heterogeneity of breast cancer independent of metastatic outcome. PAM50 classification showed a higher distribution of luminal A cells associated with good metastatic outcome. We observed a significant difference in PAM50 distributions between the individual clusters. One cluster diverges significantly from all others and shows a decrease in luminal subtype and an increase in basal like cells. Further analysis of this cluster shows an association of selected differentially expressed markers to metastatic progression. Co-occurrence and decreased expression of several EMT markers was observed with disease progression. Conclusion High tumour heterogeneity is independent of metastatic outcomes. Downregulation of EMT related genes can be essential in metastatic progression when selected markers are co-occurring in a subpopulation. Take-home message High tumour heterogeneity is present in breast cancer independent of metastatic outcomes. Downregulation of EMT related genes can be essential in metastatic progression when selected markers are co-occurring in a subpopulation.

Many signaling and other genes known as “hidden” drivers may not be genetically or epigenetically altered or differentially expressed at the mRNA or protein levels, but, rather, drive a phenotype such as tumorigenesis via post-translational modification or other mechanisms. However, conventional approaches based on genomics or differential expression are limited in exposing such hidden drivers. Here, we present a comprehensive algorithm and toolkit NetBID2 (data-driven network-based Bayesian inference of drivers, version 2), which reverse-engineers context-specific interactomes and integrates network activity inferred from large-scale multi-omics data, empowering the identification of hidden drivers that could not be detected by traditional analyses. NetBID2 has substantially re-engineered the previous prototype version by providing versatile data visualization and sophisticated statistical analyses, which strongly facilitate researchers for result interpretation through end-to-end multi-omics data analysis. We demonstrate the power of NetBID2 using three hidden driver examples. We deploy NetBID2 Viewer, Runner, and Cloud apps with 145 context-specific gene regulatory and signaling networks across normal tissues and paediatric and adult cancers to facilitate end-to-end analysis, real-time interactive visualization and cloud-based data sharing. NetBID2 is freely available at https://jyyulab.github.io/NetBID.

Major depressive disorder (MDD) is a heterogeneous and prevalent mental disorder associated with increased morbidity, disability, and mortality. However, its underlying mechanisms remain unclear. All analyses were conducted based on integrated samples from the GEO database. Differential expression analysis, unsupervised consensus clustering analysis, enrichment analysis, and regulation network analysis were performed. Mitogen-activated protein kinase (MAPK) signaling pathway was identified as an associated pathway in the development of MDD. From transcriptional signatures, we classified the MDD patients into two subgroups using unsupervised clustering and revealed 13 differential expression genes between subgroups, which indicates the probably relative complications. We further illustrated potential molecular mechanisms of MDD, including dysregulation in the neurotrophin signaling pathway, peptidyl-serine phosphorylation, and endocrine resistance. Moreover, we identified hub genes, including MAPK8, TP53, and HRAS in the maintenance of MDD. Furthermore, we demonstrated that the axis of miRNAs-TFs-HRAS/TP53/MAPK8 may play a critical role in MDD. Taken together, we demonstrated an overview of MAPK-related key genes in MDD, determined two molecular subtypes, and identified the key genes and core network that may contribute to the procession of MDD.

Purpose This study aims to identify potential prognostic biomarkers of bladder cancer (BCa) based on large-scale multi-omics data and investigate the role of SRC in improving predictive outcomes for BCa patients and those receiving immune checkpoint therapies (ICTs). Methods Large-scale multi-comic data were enrolled from the Cancer Proteome Atlas, the Cancer Genome Atlas and gene expression omnibus based on machining-learning methods. Immune infiltration, survival and other statistical analyses were implemented using R software in cancers (n = 12,452). The predictive value of SRC was performed in 81 BCa patients receiving ICT from aa validation cohort (n = 81). Results Landscape of novel candidate prognostic protein signatures of BCa patients was identified. Differential BECLIN, EGFR, PKCALPHA, ANNEXIN1, AXL and SRC expression significantly correlated with the outcomes for BCa patients from multiply cohorts (n = 906). Notably, risk score of the integrated prognosis-related proteins (IPRPs) model exhibited high diagnostic accuracy and consistent predictive ability (AUC = 0.714). Besides, we tested the clinical relevance of baseline SRC protein and mRNA expression in two independent confirmatory cohorts (n = 566) and the prognostic value in pan-cancers. Then, we found that elevated SRC expression contributed to immunosuppressive microenvironment mediated by immune checkpoint molecules of BCa and other cancers. Next, we validated SRC expression as a potential biomarker in predicting response to ICT in 81 BCa patient from FUSCC cohort, and found that expression of SRC in the baseline tumour tissues correlated with improved survival benefits, but predicts worse ICT response. Conclusion This study first performed the large-scale multi-omics analysis, distinguished the IPRPs (BECLIN, EGFR, PKCALPHA, SRC, ANNEXIN1 and AXL) and revealed novel prediction model, outperforming the currently traditional prognostic indicators for anticipating BCa progression and better clinical strategies. Additionally, this study provided insight into the importance of biomarker SRC for better prognosis, which may inversely improve predictive outcomes for patients receiving ICT and enable patient selection for future clinical treatment.

BackgroundAlzheimer’s Disease (AD) is heterogeneous and shows complex interconnected pathways at various biological levels. Risk scores contribute greatly to disease prognosis and biomarker discovery but typically represent generic risk factors. However, large-scale multi-omics data can generate individualized risk factors. Filtering these risk factors with brain-derived extracellular vesicles (EVs) could yield key pathologic pathways and vesicular vehicles for treatment delivery.MethodsA list of 460 EV-related genes was curated from brain tissue samples in the ExoCarta database. This list was used to select genes from transcriptomics, proteomics, and DNA methylation data. Significant risk factors included demographic features (age, sex) and genes significant for progression in transcriptomics data. These genes were selected using Cox regression, aided by the Least Absolute Shrinkage and Selection Operator (LASSO), and were used to construct three risk models at different omics levels. Gene signatures from the significant risk factors were used as biomarkers for further evaluation, including gene set enrichment analysis (GSEA) and drug perturbation analysis.ResultsNine EV-related genes were identified as significant risk factors. All three risk models predicted high/low risk groups with significant separation in Kaplan-Meier analysis. Training the transcriptomics risk models on EV-related genes yielded better AD classification results than using all genes in an independent dataset. GSEA revealed Mitophagy and several other significant pathways related to AD. Four drugs showed therapeutic potential to target the identified risk factors based on Connectivity Map analysis.ConclusionThe proposed risk score model demonstrates a novel approach to AD using EV-related large-scale multi-omics data. Potential biomarkers and pathways related to AD were identified for further investigation. Drug candidates were identified for further evaluation in biological experiments, potentially transported to targeted tissues via bioengineered EVs.

During the past two decades, computational analysis has become paramount for biological research. Advancements in high-throughput methods and computational tools resulted in the generation of large amounts of data from different omics fields (multi-omics), such as genomics, epigenomics, transcriptomics, and metabolomics. This plethora of large scale and diverse omics data is being driven by the understanding that a single-omic type does not provide adequate information and integrative analysis of multi-omics data is optimal to gain sufficiently meaningful insights into the actual biological mechanisms. Although various open-source tools have been developed for this purpose, multi-omics data integration and analysis are still beset by a number of problems, including software compatibility, complex parameter selection and creation of functional pipelines with multiple steps of analyses. In this work we present GenOptics, a novel visual analytics platform that aims to facilitate the integration and subsequent analysis of diverse multi-omics datasets as well as meta-data (e.g., clinical data), through a fully interactive environment. The platform comprises of two separate parts. The first incorporates asynchronous analyses of Next-Generation Sequencing raw data, including RNA-, Whole exome-, and ChIP-Seq, using workflows implemented with the Common Workflow Language (doi:10.6084/m9.figshare.3115156.v2) and Docker containers to automate software installation and confer cross-platform portability. The second part constitutes the analytical platform itself, designed to facilitate the execution of robust bioinformatics analyses by life scientists with minimal or no knowledge of programming. GenOptics constitutes an open source (https://genoptics.github.io/), computational biology platform for novel pattern and biomarker detection.

It is obvious that the large scaled data can be generated as well as processed by implementing the most effective computational techniques. In this regard, applications inked to operation management, transact generation, health care as well as industrial applications require specific trends and patterns within these large socioeconomic datasets. Sometimes, it can be a point of discussion regarding specifying the parameters associated with the voluminous data to prioritize the granular information about the individual cluster. Also in many cases, emphasis can be given to analyze the social networks and social engagement behaviors of individuals by mapping mobility patterns implementing sensors or mechanisms as well as usage of remote sensors to track all the patterns provisioning the coordination with information communication. In some cases also, based on the web analytics along with machine learning, prediction associated with large scaled data invites the opportunities to new mechanisms with conceptual applications in management sector also. While concentrating on granular data, it is essential to entrust the key sources of the voluminous data whether private, public or self quantified. So adoption of the recent mechanisms can lead to generate ambient data which can partially be emitted to be linked with dynamic networks quantifying the actions and behaviors. It is observed that the size and dimension of data while associated and shared in business and general applications are enhanced immeasurably. The textual data may be structured or unstructured. Similarly, the images and social media sites linked to multiplicity platforms can be generated in voluminous structure to be the evident to strategic technology trends. Considering this trend, partially the machine learning techniques or evolutionary as well as heuristic techniques can be applied to prioritize and focus on the majority of data to overcome the specific challenges.KeywordsBig dataHeuristicsNon-homogeneous dataParameterized costFunctional values

Receptor interactions with short linear peptide fragments (ligands) are at the base of many biological signaling processes. Conserved and information-rich amino acid patterns, commonly called sequence motifs, shape and regulate these interactions. Because of the properties of a receptor-ligand system or of the assay used to interrogate it, experimental data often contain multiple sequence motifs. GibbsCluster is a powerful tool for unsupervised motif discovery because it can simultaneously cluster and align peptide data. The GibbsCluster 2.0 presented here is an improved version incorporating insertion and deletions accounting for variations in motif length in the peptide input. In basic terms, the program takes as input a set of peptide sequences and clusters them into meaningful groups. It returns the optimal number of clusters it identified, together with the sequence alignment and sequence motif characterizing each cluster. Several parameters are available to customize cluster analysis, including adjustable penalties for small clusters and overlapping groups and a trash cluster to remove outliers. As an example application, we used the server to deconvolute multiple specificities in large-scale peptidome data generated by mass spectrometry. The server is available at http://www.cbs.dtu.dk/services/GibbsCluster-2.0.

BackgroundColon cancer is common worldwide and is the leading cause of cancer-related death. Multiple levels of omics data are available due to the development of sequencing technologies. In this study, we proposed an integrative prognostic model for colon cancer based on the integration of clinical and multi-omics data.MethodsIn total, 344 patients were included in this study. Clinical, gene expression, DNA methylation and miRNA expression data were retrieved from The Cancer Genome Atlas (TCGA). To accommodate the high dimensionality of omics data, unsupervised clustering was used as dimension reduction method. The bias-corrected Harrell’s concordance index was used to verify which clustering result provided the best prognostic performance. Finally, we proposed a prognostic prediction model based on the integration of clinical data and multi-omics data. Uno’s concordance index with cross-validation was used to compare the discriminative performance of the prognostic model constructed with different covariates.ResultsCombinations of clinical and multi-omics data can improve prognostic performance, as shown by the increase of the bias-corrected Harrell’s concordance of the prognostic model from 0.7424 (clinical features only) to 0.7604 (clinical features and three types of omics features). Additionally, 2-year, 3-year and 5-year Uno’s concordance statistics increased from 0.7329, 0.7043, and 0.7002 (clinical features only) to 0.7639, 0.7474 and 0.7597 (clinical features and three types of omics features), respectively.ConclusionIn conclusion, this study successfully combined clinical and multi-omics data for better prediction of colon cancer prognosis.

The discovery of cancer subtypes based on unsupervised clustering helps in providing a precise diagnosis, guide treatment, and improve patients’ prognoses. Instead of single-omics data, multi-omics data can improve the clustering performance because it obtains a comprehensive landscape for understanding biological systems and mechanisms. However, heterogeneous data from multiple sources raises high complexity and different kinds of noise, which are detrimental to the extraction of clustering information. We propose an end-to-end deep learning based method, called Multi-omics Clustering Variational Autoencoders (MCluster-VAEs), that can extract cluster-friendly representations on multi-omics data. First, a unified network architecture with an attention mechanism was developed for accurately modeling multi-omics data. Then, using a novel objective function built from the Variational Bayes technique, the model was trained to effectively obtain the posterior estimation of the clustering assignments. Compared with 12 other state-of-the-art multi-omics clustering methods, MCluster-VAEs achieved an outstanding performance on benchmark datasets from the TCGA database. On the Pan Cancer dataset, MCluster-VAEs achieved an adjusted Rand index of approximately 0.78 for cancer category recognition, an increase of more than 18% compared with other methods. Furthermore, a survival analysis and clinical parameter enrichment tests conducted on 10 cancer datasets demonstrated that MCluster-VAEs provides comparable and even better results than many common integrative approaches. These results demonstrate that MCluster-VAEs are a powerful new tool for dissecting complex multi-omics relationships and providing new insights for cancer subtype discovery.