Hulin Wu

University of Rochester, Rochester, New York, United States

Are you Hulin Wu?

Claim your profile

Publications (79)197.76 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: An improved filtering method is provided to estimate the parameter for a type of nonlinear multivariate stochastic differential equations (SDEs) with multiplicative noise, when discrete observations contaminated with measurement error are given. First, a transformation is used to transform the diffusion terms of the SDEs into unit diffusion such that the improved filtering method can be used. After the transformation, the drift terms of the SDEs are local linearized by means of Itô formula rather than Taylor expansion, and the predictions of the innovation estimators are approximated by a more rigorous theoretical form which guarantees that the improved method works well even when the Jacobian matrix of the drift terms is singular or ill-conditioned. The parameter is estimated from discrete observations by maximum likelihood technique. The improved method is compared to an existing software tool CTSM by estimating Van der Pol’s random oscillation with unobserved state variables, the provided method proves to be robust particularly when observation noise is relatively large. Applying the improved method to modified stochastic Lotka–Volterra equations with multiplicative noise, where the performance of the linear approximation by Itô formula and Taylor expansion is compared, in conclusion the provided method has better performance especially under long observation time interval.
    Computational Statistics & Data Analysis 11/2014; 79:113–119. · 1.30 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Respiratory epithelial cells are the primary target of influenza virus infection in human. However, the molecular mechanisms of airway epithelial cell responses to viral infection are not fully understood. Revealing genome-wide transcriptional and post-transcriptional regulatory relationships can further advance our understanding of this problem, which motivates the development of novel and more efficient computational methods to simultaneously infer the transcriptional and post-transcriptional regulatory networks.
    BMC Bioinformatics 10/2014; 15(1):336. · 3.02 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: During the earlier years of the HIV/AIDS epidemic, initial reports described sensorineural hearing loss in up to 49% of individuals with HIV/AIDS. During those years, patients commonly progressed to advanced stages of HIV disease and frequently had neurological complications. However, the abnormalities on pure-tone audiometry and brainstem-evoked responses outlined in small studies were not always consistently correlated with advanced stages of HIV/AIDS. Moreover, these studies could not exclude the confounding effect of concurrent opportunistic infections and syphilis. Additional reports also have indicated that some antiretroviral medications may be ototoxic; thus, it has been difficult to make conclusions regarding the cause of changes in hearing function in HIV-infected patients. More recently, accelerated aging has been suggested as a potential explanation for the disproportionate increase in complications of aging described in many HIV-infected patients; hence, accelerated aging-associated hearing loss may also be playing a role in these patients.
    Ear and hearing. 08/2014;
  • [Show abstract] [Hide abstract]
    ABSTRACT: Recent studies demonstrate that human blood transcriptional signatures may be used to support diagnosis and clinical decisions for acute respiratory viral infections such as influenza. In this article, we propose to use a newly developed systems biology approach for time course gene expression data to identify significant dynamically response genes and dynamic gene network responses to viral infection. We illustrate the methodological pipeline by reanalyzing the time course gene expression data from a study with healthy human subjects challenged by live influenza virus. We observed clear differences in the number of significant dynamic response genes (DRGs) between the symptomatic and asymptomatic subjects and also identified DRG signatures for symptomatic subjects with influenza infection. The 505 common DRGs shared by the symptomatic subjects have high consistency with the signature genes for predicting viral infection identified in previous works. The temporal response patterns and network response features were carefully analyzed and investigated.
    Journal of Pharmacokinetics and Pharmacodynamics 07/2014; · 1.81 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In many regression problems, the relations between the covariates and the response may be nonlinear. Motivated by the application of reconstructing a gene regulatory network, we consider a sparse high-dimensional additive model with the additive components being some known nonlinear functions with unknown parameters. To identify the subset of important covariates, we propose a new method for simultaneous variable selection and parameter estimation by iteratively combining a large-scale variable screening (the nonlinear independence screening, NLIS) and a moderate-scale model selection (the nonnegative garrote, NNG) for the nonlinear additive regressions. We have shown that the NLIS procedure possesses the sure screening property and it is able to handle problems with non-polynomial dimensionality; and for finite dimension problems, the NNG for the nonlinear additive regressions has selection consistency for the unimportant covariates and also estimation consistency for the parameter estimates of the important covariates. The proposed method is applied to simulated data and a real data example for identifying gene regulations to illustrate its numerical performance.
    Statistica Sinica 07/2014; 24(3):1365-1387. · 1.44 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The B cell response to influenza infection of the respiratory tract contributes to viral clearance and establishes profound resistance to reinfection by related viruses. Numerous studies have measured virus-specific antibody-secreting cell (ASC) frequencies in different anatomical compartments after influenza infection and provided a general picture of the kinetics of ASC formation and dispersion. However, the dynamics of ASC populations are difficult to determine experimentally and have received little attention. Here, we applied mathematical modeling to investigate the dynamics of ASC growth, death, and migration over the 2-week period following primary influenza infection in mice. Experimental data for model fitting came from high frequency measurements of virus-specific IgM, IgG, and IgA ASCs in the mediastinal lymph node (MLN), spleen, and lung. Model construction was based on a set of assumptions about ASC gain and loss from the sampled sites, and also on the directionality of ASC trafficking pathways. Most notably, modeling results suggest that differences in ASC fate and trafficking patterns reflect the site of formation and the expressed antibody class. Essentially all early IgA ASCs in the MLN migrated to spleen or lung, whereas cell death was likely the major reason for IgM and IgG ASC loss from the MLN. In contrast, the spleen contributed most of the IgM and IgG ASCs that migrated to the lung, but essentially none of the IgA ASCs. This finding points to a critical role for regional lymph nodes such as the MLN in the rapid generation of IgA ASCs that seed the lung. Results for the MLN also suggest that ASC death is a significant early feature of the B cell response. Overall, our analysis is consistent with accepted concepts in many regards, but it also indicates novel features of the B cell response to influenza that warrant further investigation.
    PLoS ONE 01/2014; 9(8):e104781. · 3.53 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The immune response to viral infection is regulated by an intricate network of many genes and their products. The reverse engineering of gene regulatory networks (GRNs) using mathematical models from time course gene expression data collected after influenza infection is key to our understanding of the mechanisms involved in controlling influenza infection within a host. A five-step pipeline: detection of temporally differentially expressed genes, clustering genes into co-expressed modules, identification of network structure, parameter estimate refinement, and functional enrichment analysis, is developed for reconstructing high-dimensional dynamic GRNs from genome-wide time course gene expression data. Applying the pipeline to the time course gene expression data from influenza-infected mouse lungs, we have identified 20 distinct temporal expression patterns in the differentially expressed genes and constructed a module-based dynamic network using a linear ODE model. Both intra-module and inter-module annotations and regulatory relationships of our inferred network show some interesting findings and are highly consistent with existing knowledge about the immune response in mice after influenza infection. The proposed method is a computationally efficient, data-driven pipeline bridging experimental data, mathematical modeling, and statistical analysis. The application to the influenza infection data elucidates the potentials of our pipeline in providing valuable insights into systematic modeling of complicated biological processes.
    PLoS ONE 01/2014; 9(5):e95276. · 3.53 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The gene regulation network (GRN) is a high-dimensional complex system, which can be represented by various mathematical or statistical models. The ordinary differential equation (ODE) model is one of the popular dynamic GRN models. High-dimensional linear ODE models have been proposed to identify GRNs, but with a limitation of the linear regulation effect assumption. In this article, we propose a sparse additive ODE (SA-ODE) model, coupled with ODE estimation methods and adaptive group least absolute shrinkage and selection operator (LASSO) techniques, to model dynamic GRNs that could flexibly deal with nonlinear regulation effects. The asymptotic properties of the proposed method are established and simulation studies are performed to validate the proposed approach. An application example for identifying the nonlinear dynamic GRN of T-cell activation is used to illustrate the usefulness of the proposed method.
    Journal of the American Statistical Association 01/2014; 109(506). · 1.83 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: To identify sources of inter-subject variation in vaccine responses, we performed high-frequency sampling of human peripheral blood cells post-vaccination, followed by a novel systems biology analysis. Functional principal component analysis was used to examine time varying B cell vaccine responses. In subjects vaccinated within the previous three years, 90% of transcriptome variation was explained by a single subject-specific mathematical pattern. Within individual vaccine response patterns, a common subset of 742 genes was strongly correlated with migrating plasma cells. Of these, 366 genes were associated with human plasmablasts differentiating in vitro. Additionally, subject-specific temporal transcriptome patterns in peripheral blood mononuclear cells identified migration of myeloid/dendritic cell lineage cells one day after vaccination. Upstream analyses of transcriptome changes suggested both shared and subject-specific transcription groups underlying larger patterns. With robust statistical methods, time-varying response characteristics of individual subjects were effectively captured along with a shared plasma cell gene signature.
    Scientific Reports 07/2013; 3:2327. · 5.08 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The ELISPOT assay is often used for cell count determination in immunological studies. Automated methods are needed to estimate cell concentrations from spot counts obtained from the assay. Three major distributions are assumed for observational cell counts. For each assumed distribution, individual least squares (LS)/ maximum likelihood and/or individual robust least squares (RLS) are applied for parameter estimation. Distributions of study endpoints (derived variables), defined as percentages of antigen-specific memory cell per total immunoglobulin G (IgG), are investigated to provide a basis for hypothesis testing. We show, under some weak conditions, that the distribution of endpoint estimates across subjects is approximately the same within a group. Thus, the t -test or the Wilcoxon Rank Sum test can be applied to detect group differences. These methods are compared through simulations and application to real data.
    Journal of Biopharmaceutical Statistics 07/2013; 23(4):921-936. · 0.73 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: With the development of micron-scale imaging techniques, capillaries can be conveniently visualized using methods such as two-photon and whole mount microscopy. However, the presence of background staining, leaky vessels and the diffusion of small fluorescent molecules can lead to significant complexity in image analysis and loss of information necessary to accurately quantify vascular metrics. One solution to this problem is the development of accurate thresholding algorithms that reliably distinguish blood vessels from surrounding tissue. Although various thresholding algorithms have been proposed, our results suggest that without appropriate pre- or post-processing, the existing approaches may fail to obtain satisfactory results for capillary images that include areas of contamination. In this study, we propose a novel local thresholding algorithm, called directional histogram ratio at random probes (DHR-RP). This method explicitly considers the geometric features of tube-like objects in conducting image binarization, and has a reliable performance in distinguishing small vessels from either clean or contaminated background. Experimental and simulation studies suggest that our DHR-RP algorithm is superior over existing thresholding methods.
    Pattern Recognition 07/2013; 46(7):1933-1948. · 2.58 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Recently, melanoma has become the most malignant and commonly occurring skin cancer. Melanoma is not only the major source (75%) of deaths related to skin cancer, but also it is hard to be treated by the conventional drugs. Recent research indicated that angiogenesis is an important factor for tumor initiation, expansion, and response to therapy. Thus, we proposed a novel multi-scale agent-based computational model that integrates the angiogenesis into tumor growth to study the response of melanoma cancer under combined drug treatment. Our multi-scale agent-based model can simulate the melanoma tumor growth with angiogenesis under combined drug treatment. The significant synergistic effects between drug Dox and drug Sunitinib demonstrated the clinical potential to interrupt the communication between melanoma cells and its related vasculatures. Also, the sensitivity analysis of the model revealed that diffusivity related to the micro-vasculatures around tumor tissues closely correlated with the spread, oscillation and destruction of the tumor. Simulation results showed that the 3D model can represent key features of melanoma growth, angiogenesis, and its related micro-environment. The model can help cancer researchers understand the melanoma developmental mechanism. Drug synergism analysis suggested that interrupting the communications between melanoma cells and the related vasculatures can significantly increase the drug efficacy against tumor cells.
    Theoretical Biology and Medical Modelling 06/2013; 10(1):41. · 1.46 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We review mathematical modeling and related statistical issues of HIV dynamics primarily in response to antiretroviral drug therapy in this article. We start from a basic model of virus infection and then review a number of more advanced models with considering, e.g., pharmacokinetic factors, adherence and drug resistance. Specifically, we illustrate how mathematical models can be developed and parameterized to understand effects of long-term treatment and different treatment strategies on disease progression. In addition, we discuss a variety of parameter estimation methods for differential equation models that are applicable to either within- or between-host viral dynamics.
    Advanced drug delivery reviews 04/2013; · 11.96 Impact Factor
  • Source
    Xing Qiu, Hulin Wu, Rui Hu
    [Show abstract] [Hide abstract]
    ABSTRACT: Background Quantile and rank normalizations are two widely used pre-processingtechniques designed to remove technological noise presented ingenomic data. Subsequent statistical analysis such as genedifferential expression analysis is usually based on normalizedexpressions. In this study, we find that these normalizationprocedures can have a profound impact on differential expressionanalysis, especially in terms of testing power.Results We conduct theoretical derivations to show that the testing power ofdifferential expression analysis based on quantile or ranknormalized gene expressions can never reach 100% with fixed samplesize no matter how strong the gene differentiation effects are.We perform extensive simulation analyses and find theresults corroborate theoretical predictions.Conclusions Our finding may explain why genes with well documentedstrong differentiation are not always detected in microarrayanalysis. It provides new insights in microarray experimental design and will helppractitioners in selecting proper normalization procedures.
    BMC Bioinformatics 04/2013; 14(1):124. · 3.02 Impact Factor
  • Source
    Shuang Wu, Hulin Wu
    [Show abstract] [Hide abstract]
    ABSTRACT: BACKGROUND: One of the fundamental problems in time course gene expression data analysis is to identify genes associatedwith a biological process or a particular stimulus of interest, like a treatment or virus infection. Most of theexisting methods for this problem are designed for data with longitudinal replicates. But in reality, many timecourse gene experiments have no replicates or only have a small number of independent replicates. RESULTS: We focus on the case without replicates and propose a new method for identifying differentially expressedgenes by incorporating the functional principal component analysis (FPCA) into a hypothesis testingframework. The data-driven eigenfunctions allow a flexible and parsimonious representation of time coursegene expression trajectories, leaving more degrees of freedom for the inference compared to that using aprespecified basis. Moreover, the information of all genes is borrowed for individual gene inferences. CONCLUSION: The proposed approach turns out to be more powerful in identifying time course differentially expressed genescompared to the existing methods. The improved performance is demonstrated through simulation studies anda real data application to the Saccharomyces cerevisiae cell cycle data.
    BMC Bioinformatics 01/2013; 14(1):6. · 3.02 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Human CD4 T cell recall responses to influenza virus are strongly biased towards Type 1 cytokines, producing IFNγ, IL-2 and TNFα. We have now examined the effector phenotypes of CD4 T cells in more detail, particularly focusing on differences between recent versus long-term, multiply-boosted responses. Peptides spanning the proteome of temporally distinct influenza viruses were distributed into pools enriched for cross-reactivity to different influenza strains, and used to stimulate antigen-specific CD4 T cells representing recent or long-term memory. In the general population, peptides unique to the long-circulating influenza A/New Caledonia/20/99 (H1N1) induced Th1-like responses biased toward the expression of IFNγ(+)TNFα(+) CD4 T cells. In contrast, peptide pools enriched for non-cross-reactive peptides of the pandemic influenza A/California/04/09 (H1N1) induced more IFNγ(-)IL-2(+)TNFα(+) T cells, similar to the IFNγ(-)IL-2(+) non-polarized, primed precursor T cells (Thpp) that are a predominant response to protein vaccination. These results were confirmed in a second study that compared samples taken before the 2009 pandemic to samples taken one month after PCR-confirmed A/California/04/09 infection. There were striking increases in influenza-specific TNFα(+), IFNγ(+), and IL-2(+) cells in the post-infection samples. Importantly, peptides enriched for non-cross-reactive A/California/04/09 specificities induced a higher proportion of Thpp-like IFNγ(-)IL-2(+)TNFα(+) CD4 T cells than peptide pools cross-reactive with previous influenza strains, which induced more Th1 (IFNγ(+)TNFα(+)) responses. These IFNγ(-)IL-2(+)TNFα(+) CD4 T cells may be an important target population for vaccination regimens, as these cells are induced upon infection, may have high proliferative potential, and may play a role in providing future effector cells during subsequent infections.
    PLoS ONE 01/2013; 8(3):e57275. · 3.53 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The strategies of structured treatment interruptions (STIs) of antiretroviral therapies have been proposed for clinical management of HIV infected patients, but clinical studies on STIs failed to achieve a consistent conclusion for this strategy. To evaluate the STI strategies, in particular, CD4(+) T cell count-guided STIs, and explain these controversial conclusions from different clinical studies, in this paper we propose to use piecewise HIV virus dynamic models to quantitatively explore the STI strategies and investigate their dynamic behaviors. Our analysis results indicate that CD4(+) T cell counts can be maintained above a safe level using the STI with a single threshold or a threshold window. Numerical simulations show that the CD4(+) T cell counts either fluctuate or approach a stable level for a patient, depending on the prescribed upper or lower threshold values. In particular, the CD4(+) T cell counts can be stabilized at a desired level if the threshold policy control is applied. The durations of drug-on and drug-off are very sensitive to the prescribed upper or lower threshold levels, which possibly explains why the on-off strategy with fixed schedule or an STI strategy with frequent switches are associated with the high rate of failure. Our findings suggest that it is critical to carefully choose the thresholds of CD4(+) T cell count and individualize the STIs for each individual patient based on initial CD4(+) T cell counts.
    Journal of Theoretical Biology 05/2012; 308:123-34. · 2.35 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Although previous studies have found minimal changes in CD4 T cell responses after vaccination of adults with trivalent inactivated influenza vaccine, daily sampling and monitoring of the proliferation marker Ki-67 have now been used to reveal that a substantial fraction of influenza-specific CD4 T cells respond to vaccination. At 4-6 days after vaccination, there is a sharp rise in the numbers of Ki-67-expressing PBMC that produce IFNγ, IL-2 and/or TNFα in vitro in response to influenza vaccine or peptide. Ki-67(+) cell numbers then decline rapidly, and 10 days after vaccination, both Ki-67(+) and overall influenza-specific cell numbers are similar to pre-vaccination levels. These results provide a tool for assessing the quality and quantity of CD4 T cell responses to different influenza vaccines, and raise the possibility that the anti-influenza T cell memory response may be qualitatively altered by vaccination, even if the overall memory cell numbers do not change significantly.
    Vaccine 04/2012; 30(31):4581-4. · 3.77 Impact Factor
  • Source
    Hulin Wu, Hongqi Xue, Arun Kumar
    [Show abstract] [Hide abstract]
    ABSTRACT: Differential equations are extensively used for modeling dynamics of physical processes in many scientific fields such as engineering, physics, and biomedical sciences. Parameter estimation of differential equation models is a challenging problem because of high computational cost and high-dimensional parameter space. In this article, we propose a novel class of methods for estimating parameters in ordinary differential equation (ODE) models, which is motivated by HIV dynamics modeling. The new methods exploit the form of numerical discretization algorithms for an ODE solver to formulate estimating equations. First, a penalized-spline approach is employed to estimate the state variables and the estimated state variables are then plugged in a discretization formula of an ODE solver to obtain the ODE parameter estimates via a regression approach. We consider three different order of discretization methods, Euler's method, trapezoidal rule, and Runge-Kutta method. A higher-order numerical algorithm reduces numerical error in the approximation of the derivative, which produces a more accurate estimate, but its computational cost is higher. To balance the computational cost and estimation accuracy, we demonstrate, via simulation studies, that the trapezoidal discretization-based estimate is the best and is recommended for practical use. The asymptotic properties for the proposed numerical discretization-based estimators are established. Comparisons between the proposed methods and existing methods show a clear benefit of the proposed methods in regards to the trade-off between computational cost and estimation accuracy. We apply the proposed methods t an HIV study to further illustrate the usefulness of the proposed approaches.
    Biometrics 02/2012; 68(2):344-52. · 1.41 Impact Factor
  • Haihong Zhu, Hulin Wu
    [Show abstract] [Hide abstract]
    ABSTRACT: In this article, we propose a penalized likelihood method to estimate time-varying parameters in standard linear state space models. The time-varying parameter is modeled as a smoothing spline and then expressed as a state space model. The maximum likelihood method is used to estimate the smoothing parameter. The proposed method is assessed by a simulation study and applied to virological response data from an HIV-infected patient receiving antiretroviral treatment.
    Journal of Computational and Graphical Statistics 01/2012; 16(4):813-832. · 1.27 Impact Factor

Publication Stats

900 Citations
197.76 Total Impact Points

Institutions

  • 2004–2014
    • University of Rochester
      • • Department of Biostatistics and Computational Biology
      • • School of Medicine and Dentistry
      • • Department of Medicine
      Rochester, New York, United States
    • University Center Rochester
      • Department of Medicine
      Rochester, Minnesota, United States
    • University of Alabama at Birmingham
      • Division of Clinical Pharmacology
      Birmingham, Alabama, United States
  • 2012
    • Shaanxi Normal University
      Xi’an, Guangdong, China
  • 2010–2011
    • New York University College of Dentistry
      New York City, New York, United States
  • 2008
    • University of South Florida
      • Department of Epidemiology and Biostatistics
      Tampa, FL, United States
    • San Diego State University
      San Diego, California, United States