Zhonghua Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Zhonghua Liu has a total of 577 co-authors. Here are the top co-authors by number of shared publications:
Minhao Yao
Centre for Quantitative Medicine, Duke-NUS Medical School, National University of Singapore, Singapore, Singapore
17 shared publications- Tao Huang
Department of Epidemiology and Biostatistics, School of Public Health, Peking University, Beijing, China
11 shared publications - Xihong Lin
Harvard University
10 shared publications Zhenhuang Zhuang
Department of Epidemiology and Biostatistics, School of Public Health, Peking University, Beijing, 100871, China
8 shared publicationsWenxiu Wang
Department of Epidemiology and Biostatistics, School of Public Health, Peking University, Beijing, 100871, China
6 shared publicationsAndrea A. Baccarelli
Office of the Dean, Harvard T.H. Chan School of Public Health, Boston, MA, USA
5 shared publications
Latest Publications
- Open Access
- 26 June 2026
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 12 (26) , eaec7614
Plasma GDF15 affects long-term dementia risk and alters neuroimmune signaling
Growth/differentiation factor–15 (GDF15) is a secreted cytokine strongly associated with dementia risk. However, the extent to which GDF15 represents a biomarker and driver of dementia risk remains unclear. Across multiple cohorts, we demonstrated that plasma GDF15 is associated with greater dementia risk over 15- to 25-year follow-up periods when measured in midlife, with stronger associations observed for vascular, compared to Alzheimer’s disease (AD), dementia. Two-sample Mendelian randomization supported plasma GDF15’s mechanistic role in AD and related dementias, while cohort studies linked it to cerebral small vessel disease, neurodegeneration, phosphorylated tau, and a cerebrospinal fluid proteomic signature indicative of neuroimmune activation. Exposure of cultured myeloid cells to recombinant GDF15 altered biological pathways that we subsequently demonstrated are predictive of dementia risk, including interferon/antiviral responses. These findings support circulating GDF15’s role as an early biomarker—particularly for vascular dementia and neuroinflammation—and identify the mechanisms by which it may drive dementia risk.
- 26 June 2026
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 12 (26) , eaec7614
Open AccessPlasma GDF15 affects long-term dementia risk and alters neuroimmune signaling
Growth/differentiation factor–15 (GDF15) is a secreted cytokine strongly associated with dementia risk. However, the extent to which GDF15 represents a biomarker and driver of dementia risk remains unclear. Across multiple cohorts, we demonstrated that plasma GDF15 is associated with greater dementia risk over 15- to 25-year follow-up periods when measured in midlife, with stronger associations observed for vascular, compared to Alzheimer’s disease (AD), dementia. Two-sample Mendelian randomization supported plasma GDF15’s mechanistic role in AD and related dementias, while cohort studies linked it to cerebral small vessel disease, neurodegeneration, phosphorylated tau, and a cerebrospinal fluid proteomic signature indicative of neuroimmune activation. Exposure of cultured myeloid cells to recombinant GDF15 altered biological pathways that we subsequently demonstrated are predictive of dementia risk, including interferon/antiviral responses. These findings support circulating GDF15’s role as an early biomarker—particularly for vascular dementia and neuroinflammation—and identify the mechanisms by which it may drive dementia risk.
- Top Cited PapersOpen Access
- 15 April 2026
- Published by Springer Nature in Nature
- Vol. 653 (8115) , 786-795
EBV strain interacts with host HLA to drive nasopharyngeal carcinoma risk
Epstein–Barr virus (EBV) infects more than 95% of adults worldwide but is associated with endemic nasopharyngeal carcinoma (NPC) specifically in southern China1,2,3,4. Here, through a stepwise host–EBV genome interaction analysis, we identify a genetic interaction between HLA-A*11:01 and the high-risk EBV variant 85841G as a key determinant of NPC risk. Individuals carrying a susceptible HLA-A background (HLA-A*11:01− or HLA-A*02:07+) and infected with the high-risk 85841G EBV form a dual-risk subgroup with substantially elevated, interaction-driven NPC risk, far exceeding the effects of host or virus alone. This dual-risk subgroup comprises 20.5% of the population and accounts for approximately 47% of NPC cases. We show that EBV 85841G encodes an EBNA3B peptide that binds to HLA-A*11:01 and elicits specific T cell responses capable of lysing EBV+ B cells transformed by 85841G-carrying strains, and is associated with reduced salivary viral load and lower NPC risk among A*11:01 carriers. Evolutionary analysis reveals that 85841G arose via ancient recombination events between northern and southern EBV and subsequently underwent clonal expansion in southern China, leading to co-enrichment of interacting host and viral risk factors that, in turn, contribute to NPC endemicity. These findings reveal a markedly stratified, interaction-driven risk architecture in NPC and highlight opportunities for precision prevention.
- 15 April 2026
- Published by Springer Nature in Nature
- Vol. 653 (8115) , 786-795
Top Cited PapersOpen AccessEBV strain interacts with host HLA to drive nasopharyngeal carcinoma risk
Epstein–Barr virus (EBV) infects more than 95% of adults worldwide but is associated with endemic nasopharyngeal carcinoma (NPC) specifically in southern China1,2,3,4. Here, through a stepwise host–EBV genome interaction analysis, we identify a genetic interaction between HLA-A*11:01 and the high-risk EBV variant 85841G as a key determinant of NPC risk. Individuals carrying a susceptible HLA-A background (HLA-A*11:01− or HLA-A*02:07+) and infected with the high-risk 85841G EBV form a dual-risk subgroup with substantially elevated, interaction-driven NPC risk, far exceeding the effects of host or virus alone. This dual-risk subgroup comprises 20.5% of the population and accounts for approximately 47% of NPC cases. We show that EBV 85841G encodes an EBNA3B peptide that binds to HLA-A*11:01 and elicits specific T cell responses capable of lysing EBV+ B cells transformed by 85841G-carrying strains, and is associated with reduced salivary viral load and lower NPC risk among A*11:01 carriers. Evolutionary analysis reveals that 85841G arose via ancient recombination events between northern and southern EBV and subsequently underwent clonal expansion in southern China, leading to co-enrichment of interacting host and viral risk factors that, in turn, contribute to NPC endemicity. These findings reveal a markedly stratified, interaction-driven risk architecture in NPC and highlight opportunities for precision prevention.
- Top Cited Papers
- 22 January 2026
- Published by Wiley in Statistics in Medicine
- Vol. 45 (1-2) , e70394
Mendelian Randomization Methods for Causal Inference: Estimands, Identification and Inference
Mendelian randomization (MR) has become an essential tool for causal inference in biomedical and public health research. By using genetic variants as instrumental variables, MR helps address unmeasured confounding and reverse causation, offering a quasi‐experimental framework to evaluate causal effects of modifiable exposures on health outcomes. Despite its promise, MR faces substantial methodological challenges, including invalid instruments, weak instrument bias, and design complexities across different data structures. In this tutorial review, we aim to provide a systematic overview of MR methods for causal inference, emphasizing clarity of causal interpretation, study design comparisons, availability of software tools, and practical guidance for applied scientists. We organize the review around causal estimands, ensuring that analyses are anchored to well‐defined causal questions. We discuss the problems of invalid and weak instruments, comparing available strategies for their detection and correction. We integrate discussions of population‐based versus family‐based MR designs, analyses based on individual‐level versus summary‐level data, and one‐sample versus two‐sample MR designs, highlighting their relative advantages and limitations. We also summarize recent methodological advances and software developments that extend MR to settings with many weak or invalid instruments and to modern high‐dimensional omics data. Real‐data applications, including UK Biobank and Alzheimer's disease proteomics studies, illustrate the use of these methods in practice. This review aims to serve as a tutorial‐style reference for both methodologists and applied scientists.
- 22 January 2026
- Published by Wiley in Statistics in Medicine
- Vol. 45 (1-2) , e70394
Top Cited PapersMendelian Randomization Methods for Causal Inference: Estimands, Identification and Inference
Mendelian randomization (MR) has become an essential tool for causal inference in biomedical and public health research. By using genetic variants as instrumental variables, MR helps address unmeasured confounding and reverse causation, offering a quasi‐experimental framework to evaluate causal effects of modifiable exposures on health outcomes. Despite its promise, MR faces substantial methodological challenges, including invalid instruments, weak instrument bias, and design complexities across different data structures. In this tutorial review, we aim to provide a systematic overview of MR methods for causal inference, emphasizing clarity of causal interpretation, study design comparisons, availability of software tools, and practical guidance for applied scientists. We organize the review around causal estimands, ensuring that analyses are anchored to well‐defined causal questions. We discuss the problems of invalid and weak instruments, comparing available strategies for their detection and correction. We integrate discussions of population‐based versus family‐based MR designs, analyses based on individual‐level versus summary‐level data, and one‐sample versus two‐sample MR designs, highlighting their relative advantages and limitations. We also summarize recent methodological advances and software developments that extend MR to settings with many weak or invalid instruments and to modern high‐dimensional omics data. Real‐data applications, including UK Biobank and Alzheimer's disease proteomics studies, illustrate the use of these methods in practice. This review aims to serve as a tutorial‐style reference for both methodologists and applied scientists.
- Open Access
- 7 January 2026
- Published by Oxford University Press (OUP) in Briefings in Bioinformatics
- Vol. 27 (1)
CoxMDS: multiple data splitting for high-dimensional mediation analysis with survival outcomes in epigenome-wide studies
Causal mediation analysis investigates whether the effect of an exposure on an outcome operates through intermediate variables known as mediators. Although progress has been made in high-dimensional mediation analysis, current methods do not reliably control the false discovery rate (FDR) in finite samples, especially when mediators are moderately to highly correlated or follow non-Gaussian distributions. These challenges frequently arise in DNA methylation studies. We introduce CoxMDS, a multiple data splitting method that uses Cox proportional hazards models to identify putative causal mediators for survival outcomes. CoxMDS ensures finite-sample FDR control even in the presence of correlated or non-Gaussian mediators. Through simulations, CoxMDS is shown to maintain FDR control and achieve higher statistical power compared with existing approaches. In applications to DNA methylation data with survival outcomes, CoxMDS identified eight CpG sites in The Cancer Genome Atlas that are consistent with the hypothesis that DNA methylation may mediate the effect of smoking on lung cancer survival, and two CpG sites in the Alzheimer’s Disease Neuroimaging Initiative that are consistent with the hypothesis that DNA methylation may mediate the effect of smoking on time to Alzheimer’s disease conversion.
- 7 January 2026
- Published by Oxford University Press (OUP) in Briefings in Bioinformatics
- Vol. 27 (1)
Open AccessCoxMDS: multiple data splitting for high-dimensional mediation analysis with survival outcomes in epigenome-wide studies
Causal mediation analysis investigates whether the effect of an exposure on an outcome operates through intermediate variables known as mediators. Although progress has been made in high-dimensional mediation analysis, current methods do not reliably control the false discovery rate (FDR) in finite samples, especially when mediators are moderately to highly correlated or follow non-Gaussian distributions. These challenges frequently arise in DNA methylation studies. We introduce CoxMDS, a multiple data splitting method that uses Cox proportional hazards models to identify putative causal mediators for survival outcomes. CoxMDS ensures finite-sample FDR control even in the presence of correlated or non-Gaussian mediators. Through simulations, CoxMDS is shown to maintain FDR control and achieve higher statistical power compared with existing approaches. In applications to DNA methylation data with survival outcomes, CoxMDS identified eight CpG sites in The Cancer Genome Atlas that are consistent with the hypothesis that DNA methylation may mediate the effect of smoking on lung cancer survival, and two CpG sites in the Alzheimer’s Disease Neuroimaging Initiative that are consistent with the hypothesis that DNA methylation may mediate the effect of smoking on time to Alzheimer’s disease conversion.
- Open Access
- 17 September 2025
- Published by Oxford University Press (OUP) in Genomics, Proteomics & Bioinformatics
- Vol. 23 (5)
Boosting the Power of Rare Variant Association Studies by Imputation Using Large-scale Sequencing Population
With the emergence of population-scale whole-genome sequencing (WGS), rare variants can be captured precisely. Studying rare variants explains part of the heritability of complex traits that is overlooked by conventional genome-wide association studies (GWASs). However, the extent to which imputed data can approximate or improve upon the power of WGS data in rare variant association studies remains unclear. Using the UK Biobank WGS data (n = 150,119) as the ground truth, we first evaluated the consistency of rare variants in the single-nucleotide polymorphism (SNP) array data imputed using TOPMed or HRC+UK10K reference panel. Imputation quality (average R2) of the TOPMed-imputed data reached 0.6 even for extremely rare variants with minor allele count ≤ 5. TOPMed-imputed data were closer to WGS data across three ethnic groups, with average Cramer’s V > 0.75. Furthermore, association tests were performed on 45 traits. At the same sample size (n = 150,119), neither imputed dataset outperformed WGS data, but the results of the TOPMed-imputed data were more consistent with those of WGS data. When the sample size was increased to 488,377, the number of significant rare variants identified from the TOPMed-imputed data increased by 27.71% for quantitative traits and by approximately 10-fold for binary traits. Finally, we meta-analyzed the association results of SNP array and WGS for lung cancer and epithelial ovarian cancer, respectively. Compared to WGS-based results, more significant variants and genes were identified. Our findings highlight that incorporating rare variants imputed using large-scale sequencing populations can boost the power of rare variant association studies when WGS has limited sample sizes.
- 17 September 2025
- Published by Oxford University Press (OUP) in Genomics, Proteomics & Bioinformatics
- Vol. 23 (5)
Open AccessBoosting the Power of Rare Variant Association Studies by Imputation Using Large-scale Sequencing Population
With the emergence of population-scale whole-genome sequencing (WGS), rare variants can be captured precisely. Studying rare variants explains part of the heritability of complex traits that is overlooked by conventional genome-wide association studies (GWASs). However, the extent to which imputed data can approximate or improve upon the power of WGS data in rare variant association studies remains unclear. Using the UK Biobank WGS data (n = 150,119) as the ground truth, we first evaluated the consistency of rare variants in the single-nucleotide polymorphism (SNP) array data imputed using TOPMed or HRC+UK10K reference panel. Imputation quality (average R2) of the TOPMed-imputed data reached 0.6 even for extremely rare variants with minor allele count ≤ 5. TOPMed-imputed data were closer to WGS data across three ethnic groups, with average Cramer’s V > 0.75. Furthermore, association tests were performed on 45 traits. At the same sample size (n = 150,119), neither imputed dataset outperformed WGS data, but the results of the TOPMed-imputed data were more consistent with those of WGS data. When the sample size was increased to 488,377, the number of significant rare variants identified from the TOPMed-imputed data increased by 27.71% for quantitative traits and by approximately 10-fold for binary traits. Finally, we meta-analyzed the association results of SNP array and WGS for lung cancer and epithelial ovarian cancer, respectively. Compared to WGS-based results, more significant variants and genes were identified. Our findings highlight that incorporating rare variants imputed using large-scale sequencing populations can boost the power of rare variant association studies when WGS has limited sample sizes.
- Open Access
- 4 August 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 7149
Virus-human chromatin interactions reorganise 3D genome and hijack KDM5B for promoting metastasis in nasopharyngeal carcinoma
Epstein-Barr virus, the first identified human DNA tumour virus, is detectable in more than 90% of nasopharyngeal carcinoma patients in endemic regions. The 3D chromosome conformation analysis reveals that virus‒host chromatin interactions induce the spatial reorganisation of loops and compartments, resulting in "enhancer infestation" and switch of "H3K27 bivalency" at EBV-interacting regions. Through this mechanism, EBV hijacks KDM5B, a gatekeeper of genome stability, and its binding targets, leading to aberrant activation of an EBVIR-enhancer-KDM5B signature. Cancer cells with this signature present increased MYC activation, DNA damage responses, and epigenetic plasticity of epithelial-immune dual features with metastatic potential. Our multicentre multiomics study confirms that this signature is the prerequisite for chromosome instability and can be utilised as a risk factor for distant metastasis. This study highlights a mechanism in which latent viral episomes can alter the host high-order epigenotype, promoting transcriptional rewiring and metastasis in NPC.
- 4 August 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 7149
Open AccessVirus-human chromatin interactions reorganise 3D genome and hijack KDM5B for promoting metastasis in nasopharyngeal carcinoma
Epstein-Barr virus, the first identified human DNA tumour virus, is detectable in more than 90% of nasopharyngeal carcinoma patients in endemic regions. The 3D chromosome conformation analysis reveals that virus‒host chromatin interactions induce the spatial reorganisation of loops and compartments, resulting in "enhancer infestation" and switch of "H3K27 bivalency" at EBV-interacting regions. Through this mechanism, EBV hijacks KDM5B, a gatekeeper of genome stability, and its binding targets, leading to aberrant activation of an EBVIR-enhancer-KDM5B signature. Cancer cells with this signature present increased MYC activation, DNA damage responses, and epigenetic plasticity of epithelial-immune dual features with metastatic potential. Our multicentre multiomics study confirms that this signature is the prerequisite for chromosome instability and can be utilised as a risk factor for distant metastasis. This study highlights a mechanism in which latent viral episomes can alter the host high-order epigenotype, promoting transcriptional rewiring and metastasis in NPC.
- 25 June 2025
- Published by Wiley in Current Protocols
- Vol. 5 (6) , e70168
An Introduction to Causal Inference Methods with Multi‐omics Data
Omics biomarkers play a pivotal role in personalized medicine by providing molecular‐level insights into the etiology of diseases, guiding precise diagnostics, and facilitating targeted therapeutic interventions. Recent advancements in omics technologies have resulted in an increasing abundance of multimodal omics data, providing unprecedented opportunities for identifying novel omics biomarkers for human diseases. Mendelian randomization (MR) is a practically useful causal inference method that uses genetic variants as instrumental variables to infer causal relationships between omics biomarkers and complex traits/diseases by removing hidden confounding bias. In this article, we first present current challenges in performing MR analysis with omics data and then describe four MR methods for analyzing multi‐omics data, including epigenomics, transcriptomics, proteomics, and metabolomics data, all executable within the R software environment. © 2025 Wiley Periodicals LLC.
- 25 June 2025
- Published by Wiley in Current Protocols
- Vol. 5 (6) , e70168
An Introduction to Causal Inference Methods with Multi‐omics Data
Omics biomarkers play a pivotal role in personalized medicine by providing molecular‐level insights into the etiology of diseases, guiding precise diagnostics, and facilitating targeted therapeutic interventions. Recent advancements in omics technologies have resulted in an increasing abundance of multimodal omics data, providing unprecedented opportunities for identifying novel omics biomarkers for human diseases. Mendelian randomization (MR) is a practically useful causal inference method that uses genetic variants as instrumental variables to infer causal relationships between omics biomarkers and complex traits/diseases by removing hidden confounding bias. In this article, we first present current challenges in performing MR analysis with omics data and then describe four MR methods for analyzing multi‐omics data, including epigenomics, transcriptomics, proteomics, and metabolomics data, all executable within the R software environment. © 2025 Wiley Periodicals LLC.
- 7 February 2025
- Published by Springer Nature in Nature Computational Science
- Vol. 5 (2) , 125-143
A statistical framework for multi-trait rare variant analysis in large-scale whole-genome sequencing studies
Large-scale whole-genome sequencing (WGS) studies have improved our understanding of the contributions of coding and noncoding rare variants to complex human traits. Leveraging association effect sizes across multiple traits in WGS rare variant association analysis can improve statistical power over single-trait analysis, and also detect pleiotropic genes and regions. Existing multi-trait methods have limited ability to perform rare variant analysis of large-scale WGS data. We propose MultiSTAAR, a statistical framework and computationally scalable analytical pipeline for functionally informed multi-trait rare variant analysis in large-scale WGS studies. MultiSTAAR accounts for relatedness, population structure and correlation among phenotypes by jointly analyzing multiple traits, and further empowers rare variant association analysis by incorporating multiple functional annotations. We applied MultiSTAAR to jointly analyze three lipid traits in 61,838 multi-ethnic samples from the Trans-Omics for Precision Medicine (TOPMed) Program. We discovered and replicated new associations with lipid traits missed by single-trait analysis.
- 7 February 2025
- Published by Springer Nature in Nature Computational Science
- Vol. 5 (2) , 125-143
A statistical framework for multi-trait rare variant analysis in large-scale whole-genome sequencing studies
Large-scale whole-genome sequencing (WGS) studies have improved our understanding of the contributions of coding and noncoding rare variants to complex human traits. Leveraging association effect sizes across multiple traits in WGS rare variant association analysis can improve statistical power over single-trait analysis, and also detect pleiotropic genes and regions. Existing multi-trait methods have limited ability to perform rare variant analysis of large-scale WGS data. We propose MultiSTAAR, a statistical framework and computationally scalable analytical pipeline for functionally informed multi-trait rare variant analysis in large-scale WGS studies. MultiSTAAR accounts for relatedness, population structure and correlation among phenotypes by jointly analyzing multiple traits, and further empowers rare variant association analysis by incorporating multiple functional annotations. We applied MultiSTAAR to jointly analyze three lipid traits in 61,838 multi-ethnic samples from the Trans-Omics for Precision Medicine (TOPMed) Program. We discovered and replicated new associations with lipid traits missed by single-trait analysis.
- Open Access
- 4 December 2024
- Published by Elsevier in Cell Genomics
- Vol. 4 (12) , 100700
Deciphering proteins in Alzheimer’s disease: A new Mendelian randomization method integrated with AlphaFold3 for 3D structure prediction
- 4 December 2024
- Published by Elsevier in Cell Genomics
- Vol. 4 (12) , 100700
Open AccessDeciphering proteins in Alzheimer’s disease: A new Mendelian randomization method integrated with AlphaFold3 for 3D structure prediction
- Open Access
DNA-SE: Towards Deep Neural-Nets Assisted Semiparametric Estimation
Semiparametric statistics play a pivotal role in a wide range of domains, including but not limited to missing data, causal inference, and transfer learning, to name a few. In many settings, semiparametric theory leads to (nearly) statistically optimal procedures that yet involve numerically solving Fredholm integral equations of the second kind. Traditional numerical methods, such as polynomial or spline approximations, are difficult to scale to multi-dimensional problems. Alternatively, statisticians may choose to approximate the original integral equations by ones with closed-form solutions, resulting in computationally more efficient, but statistically suboptimal or even incorrect procedures. To bridge this gap, we propose a novel framework by formulating the semiparametric estimation problem as a bi-level optimization problem; and then we develop a scalable algorithm called Deep Neural-Nets Assisted Semiparametric Estimation (DNA-SE) by leveraging the universal approximation property of Deep Neural-Nets (DNN) to streamline semiparametric procedures. Through extensive numerical experiments and a real data analysis, we demonstrate the numerical and statistical advantages of $\dnase$ over traditional methods. To the best of our knowledge, we are the first to bring DNN into semiparametric statistics as a numerical solver of integral equations in our proposed general framework.Open AccessDNA-SE: Towards Deep Neural-Nets Assisted Semiparametric Estimation
Semiparametric statistics play a pivotal role in a wide range of domains, including but not limited to missing data, causal inference, and transfer learning, to name a few. In many settings, semiparametric theory leads to (nearly) statistically optimal procedures that yet involve numerically solving Fredholm integral equations of the second kind. Traditional numerical methods, such as polynomial or spline approximations, are difficult to scale to multi-dimensional problems. Alternatively, statisticians may choose to approximate the original integral equations by ones with closed-form solutions, resulting in computationally more efficient, but statistically suboptimal or even incorrect procedures. To bridge this gap, we propose a novel framework by formulating the semiparametric estimation problem as a bi-level optimization problem; and then we develop a scalable algorithm called Deep Neural-Nets Assisted Semiparametric Estimation (DNA-SE) by leveraging the universal approximation property of Deep Neural-Nets (DNN) to streamline semiparametric procedures. Through extensive numerical experiments and a real data analysis, we demonstrate the numerical and statistical advantages of $\dnase$ over traditional methods. To the best of our knowledge, we are the first to bring DNN into semiparametric statistics as a numerical solver of integral equations in our proposed general framework.