A5017220650- Statistical Methods and Inference
- Metabolomics and Mass Spectrometry Studies
- Bioinformatics and Genomic Networks
- Gene expression and cancer classification
- Complex Network Analysis Techniques
- Sparse and Compressive Sensing Techniques
- Network Traffic and Congestion Control
- Control Systems and Identification
- Complex Systems and Time Series Analysis
- Gene Regulatory Network Analysis
- Advanced Queuing Theory Analysis
- Advanced Bandit Algorithms Research
- Bayesian Modeling and Causal Inference
- Electric Vehicles and Infrastructure
- Bayesian Methods and Mixture Models
- Statistical Methods and Bayesian Inference
- Advanced Statistical Methods and Models
- Monetary Policy and Economic Impact
- Advanced Wireless Network Optimization
- Cancer, Lipids, and Metabolism
- Advanced Battery Technologies Research
- Distributed Sensor Networks and Detection Algorithms
- Financial Risk and Volatility Modeling
- Network Security and Intrusion Detection
- Smart Grid Energy Management
Democritus University of Thrace
2025
University of Florida
2015-2024
University of California, Los Angeles
2002-2024
George Mason University
2024
Weatherford College
2024
University of Michigan–Ann Arbor
2012-2023
Botswana Open University
2021
California University of Pennsylvania
2020
Yale University
2020
Alexandra Hospital
2019
Gaussian graphical models explore dependence relationships between random variables, through the estimation of corresponding inverse covariance matrices. In this paper we develop an estimator for such appropriate data from several that share same variables and some structure. setting, estimating a single model would mask underlying heterogeneity, while separate each category does not take advantage common We propose method jointly estimates to different categories present in data, aiming...
Many scientific and economic problems involve the analysis of high-dimensional time series datasets. However, theoretical studies in statistics to date rely primarily on assumption independent identically distributed (i.i.d.) samples. In this work, we focus stable Gaussian processes investigate properties $\ell _1$-regularized estimates two important statistical context series: (a) stochastic regression with serially correlated errors (b) transition matrix estimation vector autoregressive...
Consensus clustering (CC) has been adopted for unsupervised class discovery in many genomic studies. It calculates how frequently two samples are grouped together repeated runs and uses the resulting pairwise "consensus rates" visual demonstration that clusters exist, comparing cluster stability estimating optimal number (K). However, sensitivity specificity of CC have not systemically assessed. Through simulations we find is able to divide randomly generated unimodal data into apparently...
Abstract Motivation: Recent advances in high-throughput omics technologies have enabled biomedical researchers to collect large-scale genomic data. As a consequence, there has been growing interest developing methods integrate such data obtain deeper insights regarding the underlying biological system. A key challenge for integrative studies is heterogeneity present different sources, which makes it difficult discern coordinated signal of from source-specific noise or extraneous effects....
Diabetes is associated with altered cellular metabolism, but how metabolism contributes to the development of diabetic complications unknown. We used BKS db/db mouse model investigate changes in carbohydrate and lipid kidney cortex, peripheral nerve, retina. A systems approach using transcriptomics, metabolomics, metabolic flux analysis identified tissue-specific differences, increased glucose fatty acid kidney, a moderate increase retina, decrease nerve. In was enhanced protein acetylation...
Abstract Motivation Recent technological advances in mass spectrometry, development of richer spectral libraries and data processing tools have enabled large scale metabolic profiling. Biological interpretation metabolomics studies heavily relies on knowledge-based that contain information about pathways. Incomplete coverage different areas metabolism lack non-canonical connections between metabolites limits the scope applications such tools. Furthermore, presence a number unknown features,...
Gut microbiota dysbiosis induces lupus-like autoimmunity in mice through altered tryptophan catabolism.
Transforming growth factor-β (TGF-β) induces epithelial-mesenchymal transition (EMT) of epithelial cells in both normal embryonic development and certain pathological contexts. Here, we show that TGF-β induced-EMT human lung cancer (A549; adenocarcinoma cells) mediates tumor cell migration invasion phenotypes. To gain insights into molecular events during EMT, employed a global stable isotope labeled profiling strategy using iTRAQ reagents, followed by 2DLC−MS/MS, which identified total 51...
The Gifi system of analyzing categorical data through nonlinear varieties classical multivariate analysis techniques is reviewed. characterized by the optimal scaling variables which implemented alternating least squares algorithms. main technique homogeneity presented, along with its extensions and generalizations leading to nonmetric principal components canonical correlation analysis. Several examples are used illustrate methods. A brief account stability issues areas applications also given.
Computer simulation often is used to study complex physical and engineering processes. Although a computer simulator can be viewed as an inexpensive way gain insight into system, it still computationally costly. Much of the recent work on design analysis experiments has focused scenarios where goal fit response surface or process optimization. In this article we develop sequential methodology for estimating contour from code. The approach uses stochastic model surrogate simulator. associated...
Abstract Although alterations in xenobiotic metabolism are considered causal the development of bladder cancer, precise mechanisms involved poorly understood. In this study, we used high-throughput mass spectrometry to measure over 2,000 compounds 58 clinical specimens, identifying 35 metabolites which exhibited significant changes cancer. This metabolic signature distinguished both normal and benign from Exploratory analyses metabolomic urine showed promise distinguishing cancer controls...
Abstract Motivation: Components of biological systems interact with each other in order to carry out vital cell functions. Such information can be used improve estimation and inference, obtain better insights into the underlying cellular mechanisms. Discovering regulatory interactions among genes is therefore an important problem biology. Whole-genome expression data over time provides opportunity determine how levels are affected by changes transcription genes, discover genes. Results: In...
Directed acyclic graphs are commonly used to represent causal relationships among random variables in graphical models. Applications of these models arise the study physical and biological systems where directed edges between nodes influence components system on each other. Estimation from observational data is computationally NP-hard. In addition, with same structure may be indistinguishable based observations alone. When exhibit a natural ordering, problem estimating reduces network. this...
Metabolomic profiling of prostate cancer (PCa) progression identified markedly elevated levels sarcosine (N-methyl glycine) in metastatic PCa and modest but significant elevation the metabolite urine. Here, we examine role key enzymes associated with metabolism progression. Consistent our earlier report, were significantly urine sediments compared to controls, a area under receiver operating characteristic curve 0.71. In addition, expression biosynthetic enzyme, glycine N-methyltransferase...
In order to increase the penetration of electric vehicles, a network fast charging stations that can provide drivers with certain level quality service (QoS) is needed. However, given strain such exert on power grid, and mobility loads represented by operating it efficiently challenging complex problem. this paper, we examine equipped an energy storage device propose scheme allocates them from as well routes customers. We three scenarios, gradually increasing their complexity. first one, all...
Studies of lipids in CKD, including ESRD, have been limited to measures conventional lipid profiles. We aimed systematically identify 17 different classes and associate the abundance thereof with alterations acylcarnitines, a metric β -oxidation, across stages CKD. From Clinical Phenotyping Resource Biobank Core (CPROBE) cohort 1235 adults, we selected panel 214 participants: 36 stage 1 or 2 99 3 61 4 18 5 Among participants, 110 were men (51.4%), 64 black (29.9%), 150 white (70.1%), mean...
BACKGROUND. In this study, we identified the lipidomic predictors of early type 2 diabetic kidney disease (DKD) progression, which are currently undefined.
Four different immunoassay and antibody microarray methods performed at four sites were used to measure the levels of a broad range proteins (N = 323 assays; 39, 88, 168, 28 assays respective sites; 237 unique analytes) in human serum plasma reference specimens distributed by Plasma Proteome Project (PPP) HUPO. The provided means (1) assess level systematic variation protein abundances associated with blood preparation (serum, citrate-anticoagulated-plasma, EDTA-anticoagulated-plasma, or...