Search

Abstract Incremental extreme learning machines (I-ELMs) can automatically determine the structure of neural networks and achieve high learning speeds. However, during the process of adding hidden nodes, unnecessary hidden nodes that have little relevance to the target may be added. Several studies have proposed methods to overcome this problem by measuring the relevance between hidden nodes and outputs and adding or removing hidden nodes accordingly. Random hidden nodes have the advantage of creating diverse patterns, but they encounter a problem in which hidden nodes that generate patterns with little or no relevance to the target can be added, thereby increasing the number of hidden nodes. Unlike in existing I-ELMs, which use random hidden nodes, we propose a compact I-ELM algorithm that initially adds linear regression nodes and subsequently applies a method to ensure that the hidden nodes have patterns differing from the existing ones. Based on benchmark data, we confirmed that the proposed method constructs a compact neural network structure with fewer hidden nodes compared to the existing I-ELM systems.

We show that the conventional income inequality indexes assess income inequality incorrectly because of three problems. The unequally distributed (UD) income-based approach solves the problems, decomposes income inequality into two kinds of departure from equality, and provides two indexes. The comprehensive assessment of income inequality requires the integration of two kinds of departure. This paper proposes the relative UD (RUD) income-based approach. The RUD income-based approach combines the cumulative distribution function and quantile function of the RUD income and produces a new index integrating two kinds of departure. We investigate the properties of the new index and demonstrate its applicability through example income distributions.

Potato virus Y (PVY) is an aphid-transmitted potyvirus that affects economically important solanaceous species. In this study, the phenomena and mechanisms following infection with PVY were investigated in tobacco (Nicotiana benthamiana). In tobacco plants, infection with a mild strain of PVY (PVYO) induced stunted growth in the first two leaves at the shoot apex starting 7 days post-infection (dpi), and mosaic symptoms began to appear on newly developing young leaves at 14 dpi. Using enzyme-linked immunosorbent assay and ultrastructure analysis, we confirmed that viral particles accumulated only in the upper developing leaves of infected plants. We analyzed reactive oxygen species (ROS) generation in leaves from the bottom to the top of the plants to investigate whether delayed symptom development in leaves was associated with a defense response to the virus. In addition, the ultrastructural analysis confirmed the increase of ATG4 and ATG8, which are autophagy markers by endoplasmic reticulum (ER) stress, and the expression of genes involved in viral RNA suppression. Overall, our results suggested that viral RNA silencing and induced autophagy may play a role in the inhibition of viral symptom development in host plants in response to PVYO infection.

Abstract The importance of perioperative respiration monitoring is highlighted by high incidences of postoperative respiratory complications unrelated to the original disease. The objectives of this pilot study were to (1) simultaneously acquire respiration rate (RR), tidal volume (TV), minute ventilation (MV), SpO2 and PetCO2 from patients in post-anesthesia care unit (PACU) and (2) identify a practical continuous respiration monitoring method by analyzing the acquired data in terms of their ability and reliability in assessing a patient’s respiratory status. Thirteen non-intubated patients completed this observational study. A portable electrical impedance tomography (EIT) device was used to acquire RREIT, TV and MV, while PetCO2, RRCap and SpO2 were measured by a Capnostream35. Hypoventilation and respiratory events, e.g., apnea and hypopnea, could be detected reliably using RREIT, TV and MV. PetCO2 and SpO2 provided the gas exchange information, but were unable to detect hypoventilation in a timely fashion. Although SpO2 was stable, the sidestream capnography using the oronasal cannula was often unstable and produced fluctuating PetCO2 values. The coefficient of determination (R2) value between RREIT and RRCap was 0.65 with a percentage error of 52.5%. Based on our results, we identified RR, TV, MV and SpO2 as a set of respiratory parameters for robust continuous respiration monitoring of non-intubated patients. Such a respiration monitor with both ventilation and gas exchange parameters would be reliable and could be useful not only for respiration monitoring, but in making PACU discharge decisions and adjusting opioid dosage on general hospital floor. Future studies are needed to evaluate the potential clinical utility of such an integrated respiration monitor.

Despite significant research on student learning performance, the literature on preschool education is limited. The learning performance of different institutes in China has not been satisfactory, and students are not given enough technological resources to help them improve their learning performance. Although Chinese preschool students are active learners, their learning performance is inadequate. As a result, this research aimed to discover how project-based learning, semi-immersive virtual reality, and mental health influence learning performance. For data collection, 800 questionnaires were distributed to high schools, and 290 valid questionnaires were considered for the data analysis using Smart PLS-SEM. According to the study's findings, project-based learning is critical for improving learning performance. This study's findings are significant because they show that mental health and semi-immersive virtual reality significantly mediate the relationship between project-based learning and learning performance. The study's findings are critical for educational institutions interested in improving student performance through project-based learning opportunities. Furthermore, because it employs an innovative technology acceptance model, this study has significant practical implications for project-based learning and student learning performance.

High-dimensional LASSO (Hi-LASSO) is a powerful feature selection tool for high-dimensional data. Our previous study showed that Hi-LASSO outperformed the other state-of-the-art LASSO methods. However, the substantial cost of bootstrapping and the lack of experiments for a parametric statistical test for feature selection have impeded to apply Hi-LASSO for practical applications. In this paper, the Python package and its Spark library are efficiently designed in a parallel manner for practice with real-world problems, as well as providing the capability of the parametric statistical tests for feature selection on high-dimensional data. We demonstrate Hi-LASSO's outperformance with various intensive experiments in a practical manner. Hi-LASSO will be efficiently and easily performed by using the packages for feature selection. Hi-LASSO packages are publicly available at https://github.com/datax-lab/Hi-LASSO under the MIT license. The packages can be easily installed by Python PIP, and additional documentation is available at https://pypi.org/project/hi-lasso and https://pypi.org/project/Hi-LASSO-spark.

Abstract Background Understanding the complex biological mechanisms of cancer patient survival using genomic and clinical data is vital, not only to develop new treatments for patients, but also to improve survival prediction. However, highly nonlinear and high-dimension, low-sample size (HDLSS) data cause computational challenges to applying conventional survival analysis. Results We propose a novel biologically interpretable pathway-based sparse deep neural network, named Cox-PASNet, which integrates high-dimensional gene expression data and clinical data on a simple neural network architecture for survival analysis. Cox-PASNet is biologically interpretable where nodes in the neural network correspond to biological genes and pathways, while capturing the nonlinear and hierarchical effects of biological pathways associated with cancer patient survival. We also propose a heuristic optimization solution to train Cox-PASNet with HDLSS data. Cox-PASNet was intensively evaluated by comparing the predictive performance of current state-of-the-art methods on glioblastoma multiforme (GBM) and ovarian serous cystadenocarcinoma (OV) cancer. In the experiments, Cox-PASNet showed out-performance, compared to the benchmarking methods. Moreover, the neural network architecture of Cox-PASNet was biologically interpreted, and several significant prognostic factors of genes and biological pathways were identified. Conclusions Cox-PASNet models biological mechanisms in the neural network by incorporating biological pathway databases and sparse coding. The neural network of Cox-PASNet can identify nonlinear and hierarchical associations of genomic and clinical data to cancer patient survival. The open-source code of Cox-PASNet in PyTorch implemented for training, evaluation, and model interpretation is available at: https://github.com/DataX-JieHao/Cox-PASNet.

High-throughput genomic technologies are leading to a paradigm shift in research of computational biology. Computational analysis with high-dimensional data and its interpretation are essential for the understanding of complex biological systems. Most biological data (e.g., gene expression and DNA sequence data) are high-dimensional, but consist of much fewer samples than predictors. Such high-dimension, low sample size (HDLSS) data often cause computational challenges in biological data analysis. A number of least absolute shrinkage and selection operator (LASSO) methods have been widely used for identifying biomarkers or prognostic factors in the field of bioinformatics. The LASSO solution has been improved through the development of the LASSO derivatives, including elastic-net, adaptive LASSO, relaxed LASSO, VISA, random LASSO, and recursive LASSO. However, there are several known limitations of the existing LASSO solutions: multicollinearity (particularly with different signs), subset size limitation, and the lack of the statistical test of significance. We propose a high-dimensional LASSO (Hi-LASSO) that theoretically improves a LASSO model providing better performance of both prediction and feature selection on extremely high-dimensional data. The Hi-LASSO alleviates bias introduced from bootstrapping, refines importance scores, improves the performance taking advantage of global oracle property, provides a statistical strategy to determine the number of bootstrapping, and allows tests of significance for feature selection with appropriate distribution. The performance of Hi-LASSO was assessed by comparing the existing state-of-the-art LASSO methods in extensive simulation experiments with multiple data settings. The Hi-LASSO was also applied for survival analysis with GBM gene expression data.

Abstract Background Predicting prognosis in patients from large-scale genomic data is a fundamentally challenging problem in genomic medicine. However, the prognosis still remains poor in many diseases. The poor prognosis may be caused by high complexity of biological systems, where multiple biological components and their hierarchical relationships are involved. Moreover, it is challenging to develop robust computational solutions with high-dimension, low-sample size data. Results In this study, we propose a Pathway-Associated Sparse Deep Neural Network (PASNet) that not only predicts patients’ prognoses but also describes complex biological processes regarding biological pathways for prognosis. PASNet models a multilayered, hierarchical biological system of genes and pathways to predict clinical outcomes by leveraging deep learning. The sparse solution of PASNet provides the capability of model interpretability that most conventional fully-connected neural networks lack. We applied PASNet for long-term survival prediction in Glioblastoma multiforme (GBM), which is a primary brain cancer that shows poor prognostic performance. The predictive performance of PASNet was evaluated with multiple cross-validation experiments. PASNet showed a higher Area Under the Curve (AUC) and F1-score than previous long-term survival prediction classifiers, and the significance of PASNet’s performance was assessed by Wilcoxon signed-rank test. Furthermore, the biological pathways, found in PASNet, were referred to as significant pathways in GBM in previous biology and medicine research. Conclusions PASNet can describe the different biological systems of clinical outcomes for prognostic prediction as well as predicting prognosis more accurately than the current state-of-the-art methods. PASNet is the first pathway-based deep neural network that represents hierarchical representations of genes and pathways and their nonlinear effects, to the best of our knowledge. Additionally, PASNet would be promising due to its flexible model representation and interpretability, embodying the strengths of deep learning. The open-source code of PASNet is available at https://github.com/DataX-JieHao/PASNet.

BACKGROUND:Myelinating Schwann cells (mSCs) form myelin in the peripheral nervous system. Because of the works by us and others, matrix metalloproteinase-9 (MMP-9) has recently emerged as an essential component of the Schwann cell signaling network during sciatic nerve regeneration. METHODOLOGY/PRINCIPAL FINDINGS:In the present study, using the genome-wide transcriptional profiling of normal and injured sciatic nerves in mice followed by extensive bioinformatics analyses of the data, we determined that an endogenous, specific MMP-9 inhibitor [tissue inhibitor of metalloproteinases (TIMP)-1] was a top up-regulated gene in the injured nerve. MMP-9 capture followed by gelatin zymography and Western blotting of the isolated samples revealed the presence of the MMP-9/TIMP-1 heterodimers and the activated MMP-9 enzyme in the injured nerve within the first 24 h post-injury. MMP-9 and TIMP-1 co-localized in mSCs. Knockout of the MMP-9 gene in mice resulted in elevated numbers of de-differentiated/immature mSCs in the damaged nerve. Our comparative studies using MMP-9 knockout and wild-type mice documented an aberrantly enhanced proliferative activity and, accordingly, an increased number of post-mitotic Schwann cells, short internodes and additional nodal abnormalities in remyelinated nerves of MMP-9 knockout mice. These data imply that during the first days post-injury MMP-9 exhibits a functionally important anti-mitogenic activity in the wild-type mice. Pharmacological inhibition of MMP activity suppressed the expression of Na(v)1.7/1.8 channels in the crushed nerves. CONCLUSION/SIGNIFICANCE:Collectively, our data established an essential role of the MMP-9/TIMP-1 axis in guiding the mSC differentiation and the molecular assembly of myelin domains in the course of the nerve repair process. Our findings of the MMP-dependent regulation of Na(v) channels, which we document here for the first time, provide a basis for therapeutic intervention in sensorimotor pathologies and pain.

Water distribution systems consistently supply high-quality water at suitable pressure and volume for human and industrial consumption. Meticulous water quality management is vital to these systems. South Korea, having established legal standards for water distribution in 1963, operates the National Auto Water Quality Monitoring System for real-time water quality monitoring and contamination warnings when levels exceed legal thresholds. The U.S. Environmental Protection Agency (EPA) points out that fixed thresholds can trigger an abundance of false-positive alarms, causing irregular hydraulic changes, and false-negative errors. This could potentially lead to a failure in detecting initial instances of pollution or micropollution that fall below the established threshold. To address this, our study developed an proactive contamination warning method for South Korea's monitoring system, utilizing long short-term memory (LSTM) for water quality prediction. We also employed ensemble empirical mode decomposition (EEMD) in feature engineering to enhance LSTM's prediction performance. Additionally, we devised an optimal water quality prediction model development methodology by comparing short- and long-term prediction performances. Our findings revealed that using EEMD for feature engineering improved the stability and reduced the prediction lag of LSTM, outperforming traditional methods. This refined approach offers a more reliable and efficient means of monitoring and managing water quality in distribution systems. [ABSTRACT FROM AUTHOR]