A5076552857- Complex Network Analysis Techniques
- Opinion Dynamics and Social Influence
- Bioinformatics and Genomic Networks
- Mental Health Research Topics
- Diffusion and Search Dynamics
- Point processes and geometric inequalities
- Gene expression and cancer classification
- Cancer Genomics and Diagnostics
- Bacteriophages and microbial interactions
- Data-Driven Disease Surveillance
- Human Mobility and Location-Based Analysis
- Functional Brain Connectivity Studies
- Cellular Automata and Applications
- Soil Geostatistics and Mapping
- Advanced Statistical Process Monitoring
- Morphological variations and asymmetry
- Advanced Clustering Algorithms Research
- Advanced Causal Inference Techniques
- Bayesian Methods and Mixture Models
- Multisensory perception and integration
- Advanced Statistical Methods and Models
- Digital Mental Health Interventions
- Advanced Statistical Modeling Techniques
- Transportation Planning and Optimization
- Mental Health and Patient Involvement
The Ohio State University
2016-2025
University of Illinois Urbana-Champaign
2015-2016
Abstract Ocean viruses are abundant and infect 20–40% of surface microbes. Infected cells, termed virocells, thus a predominant microbial state. Yet, virocells their ecosystem impacts understudied, precluding incorporation into models. Here we investigated how unrelated bacterial (phages) reprogram one host contrasting with different potential footprints. We independently infected the marine Pseudoalteromonas bacterium siphovirus PSA-HS2 podovirus PSA-HP1. Time-resolved multi-omics unveiled...
In recent years there has been an increased interest in statistical analysis of data with multiple types relations among a set entities. Such multi-relational can be represented as multi-layer graphs where the vertices represents entities and edges represent different them. For community detection graphs, we consider two random graph models, stochastic blockmodel (MLSBM) model restricted parameter space, (RMLSBM). We derive consistency results for assignments maximum likelihood estimators...
We consider the problem of estimating a consensus community structure by combining information from multiple layers multi-layer network using methods based on spectral clustering or low-rank matrix factorization. As general theme, these “intermediate fusion” involve obtaining low column rank optimizing an objective function and then columns for clustering. However, theoretical properties remain largely unexplored. In absence statistical guarantees functions, it is difficult to determine if...
Mindfulness-based therapies have been introduced as a treatment option to reduce the psychological severity of tinnitus, currently incurable chronic condition. This pilot study twelve subjects with tinnitus investigates relationship between measures both task-based and resting state functional magnetic resonance imaging (fMRI) severity, assessed Tinnitus Functional Index (TFI). MRI was measured at three time points: before, after, follow-up an 8-week long mindfulness-based cognitive therapy...
Viruses impact microbial systems through killing hosts, horizontal gene transfer, and altering cellular metabolism, consequently impacting nutrient cycles. A virus-infected cell, a "virocell," is distinct from its uninfected sister cell as the virus commandeers machinery to produce viruses rather than replicate cells. Problematically, virocell responses nutrient-limited conditions that abound in nature are poorly understood. Here we used biology approach investigate metabolic reprogramming...
To analyze data from multisubject experiments in neuroimaging studies, we develop a modeling framework for joint community detection group of related networks that can be considered as sample population networks. The proposed random effects stochastic block model facilitates the study differences and subject-specific variations structure. proposes putative mean structure, which is representative or under consideration but not structure any individual component network. Instead, memberships...
The Spatial AutoRegressive model (SAR) is commonly used in studies involving spatial and network data to estimate the or peer influence effects of covariates on response, taking into account dependence. While can be efficiently estimated with a Quasi maximum likelihood approach (QMLE), detrimental effect covariate measurement error QMLE how remedy it currently unknown. If are measured error, then may not have $\sqrt{n}$ convergence even inconsistent when node influenced by only limited...
Motivated by multi-subject experiments in neuroimaging studies, we develop a modeling framework for joint community detection group of related networks, which can be considered as sample from population networks. The proposed random effects stochastic block model facilitates the study differences and subject-specific variations structure. proposes putative mean structure is representative or under consideration but not any individual component network. Instead, memberships nodes vary each...
We present a method based on the orthogonal symmetric non-negative matrix tri-factorization of normalized Laplacian for community detection in complex networks. While exact factorization given order may not exist and is NP hard to compute, we obtain an approximate by solving optimization problem. establish connection factors obtained through basis invariant subspace estimated matrix, drawing parallel with spectral clustering. Using such clustering networks motivated analyzing block-diagonal...
Multi-layer networks are on a set of entities (nodes) with multiple types relations (edges) among them where each type relation/interaction is represented as network layer. As single layer networks, community detection an important task in multi-layer networks. A large group popular methods based optimizing quality function known the modularity score, which measure presence modules or communities Hence first step defining suitable score that appropriate for question. Here we introduce...
Abstract Motivation : Cancer is the process of accumulating genetic alterations that confer selective advantages to tumor cells. The order in which aberrations occur not arbitrary, and inferring events challenging due lack longitudinal samples from tumors. Moreover, a network model oncogenesis should capture biological facts such as distinct progression trajectories cancer subtypes patterns mutual exclusivity same pathways. In this paper, we present disjunctive Bayesian (DBN), novel...
Higher-order motif structures and multi-vertex interactions are becoming increasingly important in studies that aim to improve our understanding of functionalities evolution patterns networks. To elucidate the role higher-order community detection problems over complex networks, we introduce notion a Superimposed Stochastic Block Model (SupSBM). The model is based on random graph framework which certain or subgraphs generated through an independent hyperedge generation process, then replaced...
Networks and temporal point processes serve as fundamental building blocks for modeling complex dynamic relational data in various domains. We propose the latent space Hawkes (LSH) model, a novel generative model continuous-time networks of events, using representation nodes. events between nodes mutually exciting with baseline intensities dependent upon distances sender receiver specific effects. demonstrate that our proposed LSH can replicate many features observed real including...
We propose an Embedding Network Autoregressive Model (ENAR) for multivariate networked longitudinal data. assume the network is generated from a latent variable model, and these unobserved variables are included in structural peer effect model or time series autoregressive as additive effects. This approach takes unified view of two related problems, (1) modeling predicting data (2) causal influence estimation presence homophily finite Our strategy comprises estimating factors observed...
Introduction The advent of high throughput spatial transcriptomics (HST) has allowed for unprecedented characterization spatially distinct cell communities within a tissue sample. While wide range computational tools exist detecting in HST data, none allow the community connectivity, i.e., relative similarity cells and between found communities—an analysis task that can elucidate cellular dynamics important settings such as tumor microenvironment. Methods To address this gap, we introduce...
Relationships among teachers are known to influence their teaching-related perceptions. We study whether and how teachers' advising relationships (networks) related perceptions of satisfaction, students, over educational policies, recorded as responses a questionnaire (item responses). propose novel joint model network item (JNIRM) with correlated latent variables understand these co-varying ties. This methodology allows the analyst test interpret dependence between responses. Using JNIRM,...
We consider the problem of transferring knowledge from a source, or proxy, domain to new target for learning high-dimensional regression model with possibly different features. Recently, statistical properties homogeneous transfer have been investigated. However, most and multi-task methods assume that proxy domains same feature space, limiting their practical applicability. In applications, spaces are frequently inherently different, example, due inability measure some variables in...
In many application settings involving networks, such as messages between users of an on-line social network or transactions traders in financial markets, the observed data consist timestamped relational events, which form a continuous-time network. We propose Community Hawkes Independent Pairs (CHIP) generative model for networks. show that applying spectral clustering to aggregated adjacency matrix constructed from CHIP provides consistent community detection growing number nodes and time...
We consider the problem of estimating a consensus community structure by combining information from multiple layers multi-layer network using methods based on spectral clustering or low-rank matrix factorization. As general theme, these "intermediate fusion" involve obtaining low column rank optimizing an objective function and then columns for clustering. However, theoretical properties remain largely unexplored. In absence statistical guarantees functions, it is difficult to determine if...