Integrating ensemble systems biology feature selection and bimodal deep neural network for breast cancer prognosis prediction.
Ontology highlight
ABSTRACT:
SUBMITTER: Cheng LH
PROVIDER: S-EPMC8295302 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:The substantial development of high-throughput biotechnologies has rendered large-scale multi-omics datasets increasingly available. New challenges have emerged to process and integrate this large volume of information, often obtained from widely heterogeneous sources. Kernel methods have proven successful to handle the analysis of different types of datasets obtained on the same individuals. However, they usually suffer from a lack of interpretability since the original description of the individuals is lost due to the kernel embedding. We propose novel feature selection methods that are adapted to the kernel framework and go beyond the well-established work in supervised learning by addressing the more difficult tasks of unsupervised learning and kernel output learning. The method is expressed under the form of a non-convex optimization problem with a ℓ1 penalty, which is solved with a proximal gradient descent approach. It is tested on several systems biology datasets and shows good performances in selecting relevant and less redundant features compared to existing alternatives. It also proved relevant for identifying important governmental measures best explaining the time series of Covid-19 reproducing number evolution during the first months of 2020. The proposed feature selection method is embedded in the R package mixKernel version 0.8, published on CRAN. Installation instructions are available at http://mixkernel.clementine.wf/.
| S-EPMC8900155 | biostudies-literature
Project description:Phosphorylation of serine, threonine and tyrosine plays significant roles in cellular signal transduction and in modifying multiple protein functions. Phosphoproteins are coordinated and regulated by a network of kinases, phosphatases and phospho-binding proteins, which modify the phosphorylation states, recognize unique phosphopeptides, or target proteins for degradation. Detailed and complete information on the structure and dynamics of these networks is required to better understand fundamental mechanisms of cellular processes and diseases. High-throughput technologies have been developed to investigate phosphoproteomes in model organisms and human diseases. Among them, mass spectrometry (MS)-based technologies are the major platforms and have been widely applied, which has led to explosive growth of phosphoproteomic data in recent years. New bioinformatics tools are needed to analyze and make sense of these data. Moreover, most research has focused on individual phosphoproteins and kinases. To gain a more complete knowledge of cellular processes, systems biology approaches, including pathways and networks modeling, have to be applied to integrate all components of the phosphorylation machinery, including kinases, phosphatases, their substrates, and phospho-binding proteins. This review presents the latest developments of bioinformatics methods and attempts to apply systems biology to analyze phosphoproteomics data generated by MS-based technologies. Challenges and future directions in this field will be also discussed.
| S-EPMC4204398 | biostudies-other
Project description:MotivationA unique challenge in predictive model building for omics data has been the small number of samples (n) versus the large amount of features (p). This 'n≪p' property brings difficulties for disease outcome classification using deep learning techniques. Sparse learning by incorporating known functional relationships between the biological units, such as the graph-embedded deep feedforward network (GEDFN) model, has been a solution to this issue. However, such methods require an existing feature graph, and potential mis-specification of the feature graph can be harmful on classification and feature selection.ResultsTo address this limitation and develop a robust classification model without relying on external knowledge, we propose a forest graph-embedded deep feedforward network (forgeNet) model, to integrate the GEDFN architecture with a forest feature graph extractor, so that the feature graph can be learned in a supervised manner and specifically constructed for a given prediction task. To validate the method's capability, we experimented the forgeNet model with both synthetic and real datasets. The resulting high classification accuracy suggests that the method is a valuable addition to sparse deep learning models for omics data.Availability and implementationThe method is available at https://github.com/yunchuankong/forgeNet.Contacttianwei.yu@emory.edu.Supplementary informationSupplementary data are available at Bioinformatics online.
| S-EPMC7267822 | biostudies-literature
Project description:Emotional awareness perception is a largely growing field that allows for more natural interactions between people and machines. Electroencephalography (EEG) has emerged as a convenient way to measure and track a user's emotional state. The non-linear characteristic of the EEG signal produces a high-dimensional feature vector resulting in high computational cost. In this paper, characteristics of multiple neural networks are combined using Deep Feature Clustering (DFC) to select high-quality attributes as opposed to traditional feature selection methods. The DFC method shortens the training time on the network by omitting unusable attributes. First, Empirical Mode Decomposition (EMD) is applied as a series of frequencies to decompose the raw EEG signal. The spatiotemporal component of the decomposed EEG signal is expressed as a two-dimensional spectrogram before the feature extraction process using Analytic Wavelet Transform (AWT). Four pre-trained Deep Neural Networks (DNN) are used to extract deep features. Dimensional reduction and feature selection are achieved utilising the differential entropy-based EEG channel selection and the DFC technique, which calculates a range of vocabularies using k-means clustering. The histogram characteristic is then determined from a series of visual vocabulary items. The classification performance of the SEED, DEAP and MAHNOB datasets combined with the capabilities of DFC show that the proposed method improves the performance of emotion recognition in short processing time and is more competitive than the latest emotion recognition methods.
| S-EPMC7374326 | biostudies-literature
Project description:Systems biology is an approach to biology that emphasizes the structure and dynamic behavior of biological systems and the interactions that occur within them. To succeed, systems biology crucially depends on the accessibility and integration of data across domains and levels of granularity. Biomedical ontologies were developed to facilitate such an integration of data and are often used to annotate biosimulation models in systems biology.We provide a framework to integrate representations of in silico systems biology with those of in vivo biology as described by biomedical ontologies and demonstrate this framework using the Systems Biology Markup Language. We developed the SBML Harvester software that automatically converts annotated SBML models into OWL and we apply our software to those biosimulation models that are contained in the BioModels Database. We utilize the resulting knowledge base for complex biological queries that can bridge levels of granularity, verify models based on the biological phenomenon they represent and provide a means to establish a basic qualitative layer on which to express the semantics of biosimulation models.We establish an information flow between biomedical ontologies and biosimulation models and we demonstrate that the integration of annotated biosimulation models and biomedical ontologies enables the verification of models as well as expressive queries. Establishing a bi-directional information flow between systems biology and biomedical ontologies has the potential to enable large-scale analyses of biological systems that span levels of granularity from molecules to organisms.
| S-EPMC3170340 | biostudies-literature
Project description:Systems biology is a data-heavy field that focuses on systems-wide depictions of biological phenomena necessarily sacrificing a detailed characterization of individual components. As an example, genome-wide protein interaction networks are widely used in systems biology and continuously extended and refined as new sources of evidence become available. Despite the vast amount of information about individual protein structures and protein complexes that has accumulated in the past 50 years in the Protein Data Bank, the data, computational tools, and language of structural biology are not an integral part of systems biology. However, increasing effort has been devoted to this integration, and the related literature is reviewed here. Relationships between proteins that are detected via structural similarity offer a rich source of information not available from sequence similarity, and homology modeling can be used to leverage Protein Data Bank structures to produce 3D models for a significant fraction of many proteomes. A number of structure-informed genomic and cross-species (i.e., virus-host) interactomes will be described, and the unique information they provide will be illustrated with a number of examples. Tissue- and tumor-specific interactomes have also been developed through computational strategies that exploit patient information and through genetic interactions available from increasingly sensitive screens. Strategies to integrate structural information with these alternate data sources will be described. Finally, efforts to link protein structure space with chemical compound space offer novel sources of information in drug design, off-target identification, and the identification of targets for compounds found to be effective in phenotypic screens.
| S-EPMC8095114 | biostudies-literature
Project description:There are significant gaps in our understanding of the pathways by which drugs act. This incomplete knowledge limits our ability to use mechanistic molecular information rationally to repurpose drugs, understand their side effects, and predict their interactions with other drugs. Here, we present DrugRouter, a novel method for generating drug-specific pathways of action by linking target genes, disease genes, and pharmacogenes using gene interaction networks. We construct pathways for more than a hundred drugs and show that the genes included in our pathways (i) co-occur with the query drug in the literature, (ii) significantly overlap or are adjacent to known drug-response pathways, and (iii) are adjacent to genes that are hits in genome-wide association studies assessing drug response. Finally, these computed pathways suggest novel drug-repositioning opportunities (e.g., statins for follicular thyroid cancer), gene-side effect associations, and gene-drug interactions. Thus, DrugRouter generates hypotheses about drug actions using systems biology data.
| S-EPMC4029855 | biostudies-literature
Project description:BACKGROUND: The goal of information integration in systems biology is to combine information from a number of databases and data sets, which are obtained from both high and low throughput experiments, under one data management scheme such that the cumulative information provides greater biological insight than is possible with individual information sources considered separately. RESULTS: Here we present PathSys, a graph-based system for creating a combined database of networks of interaction for generating integrated view of biological mechanisms. We used PathSys to integrate over 14 curated and publicly contributed data sources for the budding yeast (S. cerevisiae) and Gene Ontology. A number of exploratory questions were formulated as a combination of relational and graph-based queries to the integrated database. Thus, PathSys is a general-purpose, scalable, graph-data warehouse of biological information, complete with a graph manipulation and a query language, a storage mechanism and a generic data-importing mechanism through schema-mapping. CONCLUSION: Results from several test studies demonstrate the effectiveness of the approach in retrieving biologically interesting relations between genes and proteins, the networks connecting them, and of the utility of PathSys as a scalable graph-based warehouse for interaction-network integration and a hypothesis generator system. The PathSys's client software, named BiologicalNetworks, developed for navigation and analyses of molecular networks, is available as a Java Web Start application at http://brak.sdsc.edu/pub/BiologicalNetworks.
| S-EPMC1409799 | biostudies-other
Project description:Chemicals may lead to acute liver injuries, posing a serious threat to human health. Achieving the precise safety profile of a compound is challenging due to the complex and expensive testing procedures. In silico approaches will aid in identifying the potential risk of drug candidates in the initial stage of drug development and thus mitigating the developmental cost. In current studies, QSAR models were developed for hepatotoxicity predictions using the ensemble strategy to integrate machine learning (ML) and deep learning (DL) algorithms using various molecular features. A large dataset of 2588 chemicals and drugs was randomly divided into training (80%) and test (20%) sets, followed by the training of individual base models using diverse machine learning or deep learning based on three different kinds of descriptors and fingerprints. Feature selection approaches were employed to proceed with model optimizations based on the model performance. Hybrid ensemble approaches were further utilized to determine the method with the best performance. The voting ensemble classifier emerged as the optimal model, achieving an excellent prediction accuracy of 80.26%, AUC of 82.84%, and recall of over 93% followed by bagging and stacking ensemble classifiers method. The model was further verified by an external test set, internal 10-fold cross-validation, and rigorous benchmark training, exhibiting much better reliability than the published models. The proposed ensemble model offers a dependable assessment with a good performance for the prediction regarding the risk of chemicals and drugs to induce liver damage.
| S-EPMC11373136 | biostudies-literature
Project description:The promise of quantum computing to open new unexplored possibilities in several scientific fields has been long discussed, but until recently the lack of a functional quantum computer has confined this discussion mostly to theoretical algorithmic papers. It was only in the last few years that small but functional quantum computers have become available to the broader research community. One paradigm in particular, quantum annealing, can be used to sample optimal solutions for a number of NP-hard optimization problems represented with classical operations research tools, providing an easy access to the potential of this emerging technology. One of the tasks that most naturally fits in this mathematical formulation is feature selection. In this paper, we investigate how to design a hybrid feature selection algorithm for recommender systems that leverages the domain knowledge and behavior hidden in the user interactions data. We represent the feature selection as an optimization problem and solve it on a real quantum computer, provided by D-Wave. The results indicate that the proposed approach is effective in selecting a limited set of important features and that quantum computers are becoming powerful enough to enter the wider realm of applied science.
| S-EPMC8391326 | biostudies-literature