About

Chi Wei Cliburn Chan is a Professor of Biostatistics & Bioinformatics, a Research Professor of Statistical Science, and a Research Professor of Mathematics at Duke University. He is also a member of the Duke Cancer Institute. His professional roles are based at Duke's Department of Biostatistics and Bioinformatics, located at 2424 Erwin Road, Hock Plaza. Dr. Chan's work involves research in biostatistics, bioinformatics, and related fields, contributing to the academic and scientific community through his faculty positions and research activities.

Research topics

• Biology
• Medicine
• Genetics
• Gerontology
• Immunology
• Chemistry
• Virology

Selected publications

• Effect of endotoxin and alum adjuvant vaccine on peanut allergy

  UNC Libraries · 2025-01-16

  articleOpen access
  PublisherDOI

• The pentameric complex is not required for congenital CMV transmission in seronegative rhesus macaques

  Science Translational Medicine · 2025-03-12 · 4 citations

  articleOpen access

  Congenital cytomegalovirus (cCMV) is the leading infectious cause of neonatal neurological impairment worldwide, but the viral factors enabling vertical spread across the placenta remain undetermined. The pentameric complex (PC), composed of the subunits gH/gL/UL128/UL130/UL131A, has been demonstrated to be important for entry into nonfibroblast cells in vitro. These findings link the PC to broad cell tropism and virus dissemination in vivo, denoting all subunits as potential targets for intervention strategies and vaccine development. To determine the relevance of the PC for congenital transmission in a translational nonhuman primate model, we engineered a rhesus CMV (RhCMV) mutant lacking the orthologs of UL128 and UL130, which demonstrated diminished infection of epithelial cells in vitro. However, intravenous inoculation of either CD4 + T cell–depleted or immunocompetent RhCMV-seronegative pregnant rhesus macaques (RMs) in the early second trimester with the PC-deficient mutant resulted in maternal RhCMV peak plasma viremia similar to inoculations with PC-intact RhCMV, although virus shedding in saliva and urine was limited. Infections with the PC-intact virus induced IgG responses that neutralized RhCMV entry into epithelial cells in tissue culture. These responses were reduced, but not absent, from animals infected with the PC-deficient virus, which also induced IgG responses against gH. Moreover, congenital CMV transmission was confirmed in multiple animals infected with PC-deficient virus by detecting viral DNA in the amniotic fluid, indicating that transplacental transmission in RMs is not contingent on the PC.

  PublisherOA PDFDOI

• Comparison Between Cox Proportional Hazards and Machine Learning Models for the Prognostication of Recurrence and Survival Following Liver Resection for Hepatocellular Carcinoma

  Journal of Hepato-Biliary-Pancreatic Sciences · 2025-07-20

  articleOpen access

  BACKGROUND: A robust prognostication model after liver resection for hepatocellular carcinoma (HCC) can guide clinical management. We aimed to develop a prognostication model for HCC recurrence and survival following liver resection, comparing between Cox proportional hazards (CPH) and supervised machine learning models. METHODS: We studied all patients who underwent liver resection for HCC between January 1, 2000 and October 31, 2022 at our institution. We aimed to predict recurrence-free survival following resection and identify risk categories for HCC recurrence. The CPH model and two supervised machine learning models (random survival forest [RSF] and extreme gradient boosting [XGB]) were used. Model performance was assessed with C-index, time-dependent area under curve (tdAUC) and Brier score. RESULTS: We studied 1290 patients, with 737 (57.1%) experiencing an event (HCC recurrence or death) over a median follow-up duration of 19.2 months. The CPH model had the overall best performance (C-index: 0.663, tdAUC at 6 months: 0.752; 1 year: 0.740; 2 years: 0.722; 5 years: 0.624). Using this model, patients stratified based on risk score could be discriminated between low, intermediate, and high-risk groups (p < 0.001). CONCLUSION: A CPH-derived prognostication model was effective for predicting and risk stratifying recurrence and survival following liver resection for HCC.

  PublisherOA PDFDOI

• A rhesus macaque model of congenital cytomegalovirus infection reveals a spectrum of vertical transmission outcomes

  Communications Biology · 2025-11-24 · 1 citations

  articleOpen access

  Congenital cytomegalovirus (cCMV) is the leading infectious cause of birth defects worldwide, yet immune determinants of protection to inform maternal vaccine design remain elusive due to the lack of a translational animal model. Here, we characterized the outcome of primary rhesus CMV (RhCMV) infection in pregnant, immunocompetent, RhCMV-naïve rhesus macaques. RhCMV DNA was detected in amniotic fluid and/or fetal tissues in six of 12 (50% placental transmission) dams following early second trimester gestation RhCMV inoculation. Widespread tissue dissemination dominated by one of two inoculated RhCMV strains was present in one fetus (8.3% cCMV disease). RhCMV DNA detection in the amniotic fluid was associated with elevated fetal and maternal plasma TNF-alpha and reduced maternal brain-derived neurotrophic factor and IL-10 levels. Maternal RhCMV exposure during pregnancy had a broad impact on the placenta and fetus even in the absence of congenital infection, as evidenced by ubiquitous maternal-fetal interface infection, and reduced placental efficiency and small-for-gestation age fetuses compared to control pregnancies. This model provides new insights into the complexity of CMV vertical transmission and can be used to evaluate immune and viral determinants of protection against cCMV.

  PublisherOA PDFDOI

• A nonhuman primate model mirrors human congenital cytomegalovirus infection and reveals a spectrum of vertical transmission outcomes

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-01-21

  preprintOpen access

  Abstract Congenital cytomegalovirus (cCMV) is the leading infectious cause of birth defects worldwide, yet immune determinants of protection to inform design of a maternal vaccine remain elusive. Here, we characterized the outcome of primary rhesus CMV (RhCMV) infection during pregnancy in an immune competent nonhuman primate (NHP) model. RhCMV DNA was detected in amniotic fluid and/or fetal tissues in six of 12 (50% placental transmission) CMV-naive rhesus macaque dams inoculated intravenously with RhCMV in early second trimester gestation. Widespread tissue dissemination dominated by one of two inoculated RhCMV strains was present in one fetus (8.3% cCMV disease). Placental RhCMV transmission was associated with elevated fetal and maternal plasma TNF-alpha and reduced maternal brain-derived neurotrophic factor and IL-10 levels. CMV exposure during pregnancy had a broad impact on the placenta and fetus even in the absence of congenital infection as evidenced by RhCMV infection at the maternal-fetal interface in all 12 dams, along with significantly reduced placental efficiency and fetal growth metrics compared to gestation-matched control pregnancies. This NHP model recapitulates key aspects of human cCMV and provides new insight into barriers and biomarkers of successful vertical transmission. One sentence summary: The nonhuman primate model mirrors the epidemiology of human congenital CMV (cCMV) after primary infection and reveals its transmission bottlenecks.

  PublisherOA PDFDOI

• Nonhuman primate model mirroring human congenital cytomegalovirus infection reveals a spectrum of vertical transmission outcomes

  Research Square · 2025-04-23

  preprintOpen access
  PublisherOA PDFDOI

• Polygenic Risk Scores and HLA Class II Variants are Biomarkers of Corticosteroid Response in Childhood Nephrotic Syndrome

  medRxiv · 2025-08-07 · 1 citations

  preprintOpen access

  Abstract Introduction Nephrotic syndrome (NS), a common glomerular disease in children, is classified based on response to corticosteroid therapy as either steroid-sensitive nephrotic syndrome (SSNS), or steroid-resistant nephrotic syndrome (SRNS). However, there are currently no reliable predictors of therapy response at initial clinical presentation. Methods We conducted genome-wide association studies, developed polygenic risk scores (PRS) for therapy response and analyzed classical HLA alleles in 1,997 (994 discovery and 1,003 replication/validation cohorts) previously unstudied children with NS and 3,558 ancestry-matched controls. Results A significant association with HLA loci defined by variants in HLA-DQB1, HLA-DRB1 , and HLA-DQA1 were found for SSNS (but not SRNS), along with a second immune-related SSNS locus: CLEC16A . A PRS that discriminates between SSNS and SRNS was validated in two independent cohorts. The HLA haplotype HLA-DRB1*07:01∼DQA1*02:01∼DQB1*02:02 was associated with ∼4 times the risk of developing SSNS. A model incorporating HLA haplotype, PRS score, and age at onset of the disease was the best predictor of steroid responsiveness with an AUC of 0.68-0.70 and an overall classification accuracy of SSNS versus SRNS of 67-71%. Conclusions Our findings confirm that SSNS (unlike SRNS) is an immune-mediated HLA-associated disorder. The PRS for therapy response and HLA haplotype can serve as biomarkers and provide a foundation for more accurate diagnoses and tailored and individualized treatment. Translational statement To identify biomarkers of pattern of steroid responsiveness in childhood-onset nephrotic syndrome, we carried out a case-control study that included over 4,000 samples, including 994 patients with NS in the discovery phase and an additional 1,003 cases from two independent replication cohorts. We identified significant risk of steroid-sensitive NS (SSNS) at HLA class II genes and CLEC16A (a gene important in the regulation of T and B lymphocytes). The HLA haplotype DRB1*07:01∼DQA1*02:01∼DQB1*02:02 was associated with four-fold increased odds of SSNS. A model incorporating HLA haplotype, polygenic risk score, and age at onset of the disease was the best predictor of steroid responsiveness providing useful delineation of steroid sensitivity from steroid resistance. In conclusion, age at onset of disease, HLA class II variants and polygenic risk scores are useful biomarkers of corticosteroid response in childhood NS and may serve as useful clinical decision support tools to guide treatment. Abstract Figure Graphical Abstract

  PublisherOA PDFDOI

• Identifying correlates of viral rebound timing and viral control in SHIV-infected infant macaques after ART interruption

  Science Translational Medicine · 2025-10-29 · 2 citations

  article

  Evaluation of HIV cure strategies requires antiretroviral therapy (ART) interruption, but ethical and clinical considerations make this difficult in children. Here, we used a pediatric preclinical model of simian-HIV (SHIV) infection to uncover features associated with time to viral rebound (TTR) and posttreatment control (PTC) of rebound viremia after ART interruption to inform the design of studies testing cure interventions. We assessed 141 variables in SHIV-infected infant rhesus macaques in three staggered ART initiation groups and during subsequent analytical treatment interruption (ATI). Viral rebound occurred in 25 of 30 macaques within 7 to 98 days of ATI, with TTR delayed in the early ART group compared with the intermediate and late ART groups. Peak plasma viral load pre-ART was the most important correlate of TTR, with increased model performance through the successive inclusion of six additional viral and immune variables. The odds of PTC were reduced with higher SHIV RNA in rectal CD4 + T cells pre-ATI; conversely, higher frequencies of Ki67 + effector memory CD8 + T cells in lymph nodes increased the likelihood of PTC. RNA sequencing of CD4 + T cells pre-ATI revealed a down-regulated metabolic and immune gene signature in macaques with PTC. Analysis of the early ART group alone implicated transforming growth factor–β signaling genes in lack of viral rebound off ART. This comprehensive investigation reveals major determinants of viral rebound dynamics after ART interruption that should be validated and explored as potential biomarkers to screen children with HIV being considered for ATI.

  PublisherDOI

• B-Lightning: using bait genes for marker gene hunting in single-cell data with complex heterogeneity

  Briefings in Bioinformatics · 2024-11-22

  articleOpen access

  In single-cell studies, cells can be characterized with multiple sources of heterogeneity (SOH) such as cell type, developmental stage, cell cycle phase, activation state, and so on. In some studies, many nuisance SOH are of no interest, but may confound the identification of the SOH of interest, and thus affect the accurate annotate the corresponding cell subpopulations. In this paper, we develop B-Lightning, a novel and robust method designed to identify marker genes and cell subpopulations corresponding to an SOH (e.g. cell activation status), isolating it from other SOH (e.g. cell type, cell cycle phase). B-Lightning uses an iterative approach to enrich a small set of trustworthy marker genes to more reliable marker genes and boost the signals of the SOH of interest. Multiple numerical and experimental studies showed that B-Lightning outperforms existing methods in terms of sensitivity and robustness in identifying marker genes. Moreover, it increases the power to differentiate cell subpopulations of interest from other heterogeneous cohorts. B-Lightning successfully identified new senescence markers in ciliated cells from human idiopathic pulmonary fibrosis lung tissues, new T-cell memory and effector markers in the context of SARS-COV-2 infections, and their synchronized patterns that were previously neglected, new AD markers that can better differentiate AD severity, and new dendritic cell functioning markers with differential transcriptomics profiles across breast cancer subtypes. This paper highlights B-Lightning's potential as a powerful tool for single-cell data analysis, particularly in complex data sets where SOH of interest are entangled with numerous nuisance factors.

  PublisherOA PDFDOI

• A Tree Perspective on Stick-Breaking Models in Covariate-Dependent Mixtures (with Discussion)

  Bayesian Analysis · 2024-09-11 · 1 citations

  articleOpen access

  Stick-breaking (SB) processes are often adopted in Bayesian mixture models for generating mixing weights. When covariates influence the sizes of clusters, SB mixtures are particularly convenient as they can leverage their connection to binary regression to ease both the specification of covariate effects and posterior computation. Existing SB models are typically constructed based on continually breaking a single remaining piece of the unit stick. We view this from a dyadic tree perspective in terms of a lopsided bifurcating tree that extends only on one side. We show that two unsavory characteristics of SB models are in fact largely due to this lopsided tree structure. We consider a generalized class of SB models with alternative bifurcating tree structures and examine the influence of the underlying tree topology on the resulting Bayesian analysis in terms of prior assumptions, posterior uncertainty, and computational effectiveness. In particular, we provide evidence that a balanced tree topology, which corresponds to continually breaking all remaining pieces of the unit stick, can resolve or mitigate these undesirable properties of SB models that rely on a lopsided tree.

  PublisherDOI

Recent grants

• Immune defense of cCMV at the maternal-fetal interface

  NIH · $46.2M · 2019–2029

• NIH Grant RC1AI086032

  NIH · $478k · 2012

• NIH Grant RC1AI08603201

  NIH · $499k · 2011

• Clinical Core

  NIH · $69.2M · 2005–2030

• Quantitative Methods for HIV/AIDS Research

  NIH · $2.4M · 2018–2028

Frequent coauthors

• Sallie R. Permar

  Cornell University

  122 shared

• Richard Barfield

  Duke University

  70 shared

• Hsuan-Yuan Wang

  45 shared

• Kent J. Weinhold

  Duke University

  39 shared

• Janet Staats

  36 shared

• Ann Chahroudi

  Emory University

  33 shared

• Cody S. Nelson

  Brigham and Women's Hospital

  30 shared

• Justin Pollara

  Duke University

  27 shared