Search

Additional file 1: Figure S1. Flowchart highlighting different number of MiRNAs, PCGs, Metabolites analyzed in our dataset and steps taken to identify key MiRNAs, PCGs, Metabolites involved in significantly enriched metabolic pathways between OBM vs OBO. Figure S2. Correlation between clinical traits and enriched metabolic pathways. Here each block represents correlation between the enrichment profile of a metabolic pathway and a clinical characteristic based on their values across the 39 participant profiles. Here we only considered those pathways which were significantly enriched and those clinical traits which were highlighted in Fig. 3A. Figure S3. Gene expression of 25 identified genes (differentially expressed in the samples) across multiple tissue types in the GTEx portal. ( https://gtexportal.org/ ). The expression of these genes is higher across multiple tissue types and not biased towards blood tissue from which samples were collected.

Accurately measuring sleep and its quality with polysomnography (PSG) is an expensive task. Actigraphy, an alternative, has been proven cheap and relatively accurate. However, the largest experiments conducted to date, have had only hundreds of participants. In this work, we processed the data of the recently published Multi-Ethnic Study of Atherosclerosis (MESA) Sleep study to have both PSG and actigraphy data synchronized. We propose the adoption of this publicly available large dataset, which is at least one order of magnitude larger than any other dataset, to systematically compare existing methods for the detection of sleep-wake stages, thus fostering the creation of new algorithms. We also implemented and compared state-of-the-art methods to score sleep-wake stages, which range from the widely used traditional algorithms to recent machine learning approaches. We identified among the traditional algorithms, two approaches that perform better than the algorithm implemented by the actigraphy device used in the MESA Sleep experiments. The performance, in regards to accuracy and F1 score of the machine learning algorithms, was also superior to the device’s native algorithm and comparable to human annotation. Future research in developing new sleep-wake scoring algorithms, in particular, machine learning approaches, will be highly facilitated by the cohort used here. We exemplify this potential by showing that two particular deep-learning architectures, CNN and LSTM, among the many recently created, can achieve accuracy scores significantly higher than other methods for the same tasks.Other Information Published in: npj Digital Medicine License: https://creativecommons.org/licenses/by/4.0See article on publisher's website: http://dx.doi.org/10.1038/s41746-019-0126-9

Additional file 9: Correlation of 3 ir-lncRNA signature with immune subpopulations across tumor types. Pearson correlation heatmap between immune cell subpopulation enrichment scores and 3 ir-lncRNA scores.

Additional file 6: Prognostic value of ICR classifier across solid cancers. Forest plot showing HRs for death (overall survival) and corresponding 95%-confidence interval for the continuous ICR score and number of patients for each TCGA cancer cohort and RAQA breast cancer cohort. Significant positive HRs are visualized in blue and significant negative HRs are visualized in red. ICR enabled (HR < 1, p-value < 0.05) cancer types are indicated with orange asterisks and ICR disabled (HR > 1, p-value < 0.05) cancers are indicated with purple asterisks.

Breast cancer largely dominates the global cancer burden statistics; however, there are striking disparities in mortality rates across countries. While socioeconomic factors contribute to population-based differences in mortality, they do not fully explain disparity among women of African ancestry (AA) and Arab ancestry (ArA) compared to women of European ancestry (EA). In this study, we sought to identify molecular differences that could provide insight into the biology of ancestry-associated disparities in clinical outcomes. We applied a unique approach that combines the use of curated survival data from The Cancer Genome Atlas (TCGA) Pan-Cancer clinical data resource, improved single-nucleotide polymorphism-based inferred ancestry assignment, and a novel breast cancer subtype classification to interrogate the TCGA and a local Arab breast cancer dataset. We observed an enrichment of BasalMyo tumors in AA patients (38 vs 16.5% in EA, p = 1.30E − 10), associated with a significant worse overall (hazard ratio (HR) = 2.39, p = 0.02) and disease-specific survival (HR = 2.57, p = 0.03). Gene set enrichment analysis of BasalMyo AA and EA samples revealed differences in the abundance of T-regulatory and T-helper type 2 cells, and enrichment of cancer-related pathways with prognostic implications (AA: PI3K-Akt-mTOR and ErbB signaling; EA: EGF, estrogen-dependent and DNA repair signaling). Strikingly, AMPK signaling was associated with opposing prognostic connotation (AA: 10-year HR = 2.79, EA: 10-year HR = 0.34). Analysis of ArA patients suggests enrichment of BasalMyo tumors with a trend for differential enrichment of T-regulatory cells and AMPK signaling. Together, our findings suggest that the disparity in the clinical outcome of AA breast cancer patients is likely related to differences in cancer-related and microenvironmental features.Other Information Published in: npj Breast Cancer License: https://creativecommons.org/licenses/by/4.0See article on publisher's website: http://dx.doi.org/10.1038/s41523-021-00215-x

Additional file 8: Survival curves of 3 ir-lncRNA signature. Overall survival Kaplan-Meier curves in which the 3 ir-lncRNAs signature did not show any significant prognostic value. Dichotomization cutoff of ‘high’ (red) and ‘low’ (cyan) subgroups was based on the optimal cut-off point as determined by a 5-fold cross-validation analysis. Censor points are indicated by vertical lines. P-values were determined by logrank test.

Additional file 1: Mapping of differentially expressed ir-lncRNAs to protein coding genes. (A) Diagram representation of the random walk with restart global propagation network algorithm. (B) Walkscore distribution of protein-coding genes in TCGA-BRCA, with cutoff set at walkscore ≥ 0.01 to generate a ranked list of protein-coding genes in proximity of differentially expressed ir-lncRNAs.

We introduce Permutation and Structured Perturbation Inference (PSPI), a new problem formulation that abstracts many graph matching tasks that arise in systems biology. PSPI can be viewed as a robust formulation of the permutation inference or graph matching, where the objective is to find a permutation between two graphs under the assumption that a set of edges may have undergone a perturbation due to an underlying cause. For example, suppose there are two gene regulatory networks X and Y from a diseased and normal tissue respectively. Then, the PSPI problem can be used to detect if there has been a structural change between the two networks which can serve as a signature of the disease. Besides the new problem formulation, we propose an ADMM algorithm (STEPD) to solve a relaxed version of the PSPI problem. An extensive case study on comparative gene regulatory networks (GRNs) is used to demonstrate that STEPD is able to accurately infer structured perturbations and thus provides a tool for computational biologists to identify novel prognostic signatures. A spectral analysis confirms that STEPD can recover small clique-like perturbations making it a useful tool for detecting permutation-invariant changes in graphs., Comment: 8 pages, 4 Figures, 2 Tables

Motivation: A global effort is underway to identify drugs for the treatment of COVID-19. Since de novo drug design is an extremely long, time-consuming, and expensive process, efforts are underway to discover existing drugs that can berepurposed for COVID-19.Model: We propose a machine learning representation framework that uses deep learning induced vector embeddings of drugs and viral proteins as features to predict drug-viral protein activity. The prediction model in-turn is used to build an ensemble framework to rank approved drugs based on their ability to inhibit the three main proteases (enzymes) of the SARS-COV-2 virus.Results: We identify a ranked list of 19 drugs as potential targets including 7 antivirals, 6 anticancer, 3 antibiotics, 2 antimalarial, and 1 antifungal. Several drugs, such as Remdesivir, Lopinavir, Ritonavir, and Hydroxychloroquine, in our ranked list, are currently in clinical trials. Moreover, through molecular docking simulations, we demonstrate that majority of the anticancer and antibiotic drugs in our ranked list have low binding energies and thus high binding affinity with the 3CL-pro protease of SARS-COV-2 virus.Disclaimer: Our models are computational and the drugs suggested should not be taken for treating COVID-19 without a doctor's advice, as further wet-lab research and clinical trials are essential to elucidate their efficacy for this purpose.

Many bioinformatics methods have been proposed for reducing the complexity of large gene or protein networks into relevant subnetworks or modules. Yet, how such methods compare to each other in terms of their ability to identify disease-relevant modules in different types of network remains poorly understood. We launched the ‘Disease Module Identification DREAM Challenge’, an open competition to comprehensively assess module identification methods across diverse protein–protein interaction, signaling, gene co-expression, homology and cancer-gene networks. Predicted network modules were tested for association with complex traits and diseases using a unique collection of 180 genome-wide association studies. Our robust assessment of 75 module identification methods reveals top-performing algorithms, which recover complementary trait-associated modules. We find that most of these modules correspond to core disease-relevant pathways, which often comprise therapeutic targets. This community challenge establishes biologically interpretable benchmarks, tools and guidelines for molecular network analysis to study human disease biology., In this DREAM challenge, 75 methods for the identification of disease-relevant modules from molecular networks are compared and validated with GWAS data. The authors provide practical guidelines for users and establish benchmarks for network analysis.

Identification of modules in molecular networks is at the core of many current analysis methods in biomedical research. However, how well different approaches identify disease-relevant modules in different types of gene and protein networks remains poorly understood. We launched the “Disease Module Identification DREAM Challenge”, an open competition to comprehensively assess module identification methods across diverse protein-protein interaction, signaling, gene co-expression, homology, and cancer-gene networks. Predicted network modules were tested for association with complex traits and diseases using a unique collection of 180 genome-wide association studies (GWAS). Our critical assessment of 75 contributed module identification methods reveals novel top-performing algorithms, which recover complementary trait-associated modules. We find that most of these modules correspond to core disease-relevant pathways, which often comprise therapeutic targets and correctly prioritize candidate disease genes. This community challenge establishes benchmarks, tools and guidelines for molecular network analysis to study human disease biology (https://synapse.org/modulechallenge).

Fourier Transform Infrared Spectroscopy (FTIR) is a non-destructive analytical technique that has been employed in this research to characterize the biochemical make-up of various rat brain regions. The sensorimotor cortex, caudate putamen, thalamus, and the hippocampus were found to have higher olefinic content&mdash, an indicator of a higher degree of unsaturated fatty acids&mdash, rich in short-chain fatty acids, and low in ester and lipid contents. While the regions of the corpus callosum, internal, and external capsule were found to contain long-chained and higher-esterified saturated fatty acids. These molecular differences may reflect the roles of the specific regions in information processing and can provide a unique biochemical platform for future studies on the earlier detection of pathology development in the brain, as a consequence of disease or injury. Laser Ablation Inductively Coupled Plasma Mass Spectroscopy (LA-ICP-MS) is another vital analytical technique that was used in this work to analyze the elements&rsquo, distribution patterns in various regions of the brain. The complementary data sets allowed the characterization of the brain regions, the chemical dominating groups, and the elemental composition. This set-up may be used for the investigation of changes in the brain caused by diseases and help create a deeper understanding of the interactions between the organic and elemental composition.

Spectral clustering represents a successful approach to data clustering. Despite its high performance in solving complex tasks, it is often disregarded in favor of the less accurate k-means algorithm because of its computational inefficiency. In this article we present a fast in-memory spectral clustering algorithm, which can handle millions of datapoints at a desktop PC scale. The proposed technique relies on a kernel-based formulation of the spectral clustering problem, also known as kernel spectral clustering. In particular, we use a fixed-size approach based on an approximation of the feature map via the Nyström method to solve the primal optimization problem. We experimented on several small and large scale real-world datasets to show the computational efficiency and clustering quality of the proposed algorithm. ispartof: pages:557-562 ispartof: Proc. of the 24th european symposium on artificial neural networks, computational intelligence and machine learning pages:557-562 ispartof: ESANN 2016 location:Brugge, Belgium date:Apr - Apr 2016 status: published

In the modern era with the advent of technology and its widespread usage there is a huge proliferation of data. Gigabytes of data from mobile devices, market basket, geo-spatial images, search engines, online social networks etc. can be easily obtained, accumulated and stored. This immense wealth of data has resulted in massive datasets and has led to the emergence of the concept of Big Data. Mining useful information from this big data is a challenging task. With the availability of more data the choices in selecting a predictive model decreases, because very few tools arenbsp;feasible for processing large scale datasets. A successful learning framework to perform various learning tasks like classification, regression, clustering, dimensionality reduction, feature selection etc. is offered by Least Squares Support Vector Machines (LSSVM) which is designed in a primal-dual optimization setting. It provides the flexibility to extend core models by adding additional constraints to the primal problem, by changing the objective function ornbsp;introducing new model selection criteria. The goal of this thesis is to explore the role of sparsity in large scale kernel models using core models adopted from the LSSVM framework. Real-world data is often noisy and only a small fraction of it contains the most relevant information. Sparsity plays a big role in selection of this representative subset of data. We first explored sparsity in the case of large scale LSSVM using fixed-size methods with a re-weighted L1 penalty on top resulting in very sparse LSSVM (VS-LSSVM). An important aspect of kernel based methods is the selection of a subset on which the model is built and validated. We proposed a novel fast and unique representative subset (FURS) selection technique to select a subset from complex networks which retains the inherent community structure in the network. We extend this method for Big Data learning by constructing k-NN graphs out of dense data using a distributed computing platform i.e. Hadoop and then apply the FURS selection technique to obtain representative subsets on top of which models are built by kernel based methods. We then focused on scaling the kernel spectralnbsp;(KSC) technique for big data networks. We devised two model selection techniques namely balanced angular fitting (BAF) and self-tuned KSC (ST-KSC) by exploiting the structure of the projections in the eigenspace to obtain the optimal number of communities k in the large graph. A multilevel hierarchical kernel spectral clustering (MH-KSC) technique was then proposed which performs agglomerative hierarchical clustering using similarity information between the out-of-sample eigen-projections. Furthermore, we developed an algorithm to identify intervals for hierarchical clustering using the Gershgorin Circle theorem. These intervals were used to identify the optimal number of clusters at a given level of hierarchy in combination with KSC model. The MH-KSC technique was extended from networks to images and datasets using the BAF model selection criterion. We also proposed optimal sparse reductions to KSC model by reconstructing the model using a reduced set. We exploited the Group Lasso and convex re-weighted L1 penalty to sparsify the KSC model. Finally, we explored the role of re-weighted L1 penalty in case of feature selection in combination with LSSVM. We proposed a visualization (Netgram) toolkit to track the evolution of communities/clusters over time in case of dynamic time-evolving communities and datasets. Real world applications considered in this thesis include classification and regression of large scale datasets, image segmentation, flat and hierarchical community detection in large scale graphs and visualization of evolving communities. nrpages: 238 status: published

In this chapter we review the main literature related to kernel spectral clustering (KSC), an approach to clustering cast within a kernel-based optimization setting. KSC represents a least-squares support vector machine based formulation of spectral clustering described by a weighted kernel PCA objective. Just as in the classifier case, the binary clustering model is expressed by a hyperplane in a high dimensional space induced by a kernel. In addition, the multi-way clustering can be obtained by combining a set of binary decision functions via an Error Correcting Output Codes (ECOC) encoding scheme. Because of its model-based nature, the KSC method encompasses three main steps: training, validation, testing. In the validation stage model selection is performed to obtain tuning parameters, like the number of clusters present in the data. This is a major advantage compared to classical spectral clustering where the determination of the clustering parameters is unclear and relies on heuristics. Once a KSC model is trained on a small subset of the entire data, it is able to generalize well to unseen test points. Beyond the basic formulation, sparse KSC algorithms based on the Incomplete Cholesky Decomposition (ICD) and $L_0$, $L_1, L_0 + L_1$, Group Lasso regularization are reviewed. In that respect, we show how it is possible to handle large scale data. Also, two possible ways to perform hierarchical clustering and a soft clustering method are presented. Finally, real-world applications such as image segmentation, power load time-series clustering, document clustering and big data learning are considered., chapter contribution to the book "Unsupervised Learning Algorithms"

International audience

Nature methods 16(9), 843-852 (2019). doi:10.1038/s41592-019-0509-5, Published by Nature Publishing Group, London [u.a.]