Yucheng T. Yang
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
Yucheng T. Yang has a total of 280 co-authors. Here are the top co-authors by number of shared publications:
- Xing-Ming Zhao
Fudan University
13 shared publications - Zhi John Lu
School of Life Sciences, Tsinghua University, Beijing, China
8 shared publications - Anyi Yang
Department of Neurology, Zhongshan Hospital and Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China
6 shared publications - Mark Gerstein
Program in Computational Biology and Bioinformatics, Yale University, New Haven, CT 06520, USA
5 shared publications Xingzhong Zhao
Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, 200433, China
5 shared publications- Xingzhong Zhao
McWilliams School of Biomedical Informatics, University of Texas Health Science Center at Houston, Houston, USA
5 shared publications
Latest Publications
- Open Access
- 17 June 2026
- Published by Public Library of Science (PLoS) in PLOS Computational Biology
- Vol. 22 (6) , e1014422
Deciphering cell type-specific causal genetic effects on brain imaging-derived phenotypes and disorders with single-cell Mendelian randomization
Reconstructing causality routes from genetic effects to complex phenotypes in particular cell types is crucial for understanding biological mechanisms underlying the brain-associated phenotypes including imaging-derived phenotypes (IDPs), and brain disorders and behaviors (DBs). Here, we develop a single-cell Mendelian randomization framework to infer cell type-specific causal relationships between gene expression and diverse brain-associated complex phenotypes by integrating single-cell expression quantitative trait loci ( cis -eQTLs) and genome-wide association study findings. We identifiy a set of 254 and 217 cis -eQTL target genes (eGenes) that may have causal effects on 112 IDPs and 26 DBs in eight cell types, respectively. These causal eGenes exhibit strong cell type specificity and varied pleiotropy among different types of brain-associated phenotypes. Further integrative analysis reveals putative causality routes among cell type-specific causal eGenes and brain-associated complex phenotypes. Finally, we characterize the spatiotemporal expression patterns of these causal eGenes, and highlight the coordinated associations of the brain-associated phenotypes based on the expression of their causal eGenes. Overall, our study presents a large-scale analysis of the genetic effects of brain structures, disorders and behaviors, providing a catalog of cell type-specific causal eGenes.
- 17 June 2026
- Published by Public Library of Science (PLoS) in PLOS Computational Biology
- Vol. 22 (6) , e1014422
Open AccessDeciphering cell type-specific causal genetic effects on brain imaging-derived phenotypes and disorders with single-cell Mendelian randomization
Reconstructing causality routes from genetic effects to complex phenotypes in particular cell types is crucial for understanding biological mechanisms underlying the brain-associated phenotypes including imaging-derived phenotypes (IDPs), and brain disorders and behaviors (DBs). Here, we develop a single-cell Mendelian randomization framework to infer cell type-specific causal relationships between gene expression and diverse brain-associated complex phenotypes by integrating single-cell expression quantitative trait loci ( cis -eQTLs) and genome-wide association study findings. We identifiy a set of 254 and 217 cis -eQTL target genes (eGenes) that may have causal effects on 112 IDPs and 26 DBs in eight cell types, respectively. These causal eGenes exhibit strong cell type specificity and varied pleiotropy among different types of brain-associated phenotypes. Further integrative analysis reveals putative causality routes among cell type-specific causal eGenes and brain-associated complex phenotypes. Finally, we characterize the spatiotemporal expression patterns of these causal eGenes, and highlight the coordinated associations of the brain-associated phenotypes based on the expression of their causal eGenes. Overall, our study presents a large-scale analysis of the genetic effects of brain structures, disorders and behaviors, providing a catalog of cell type-specific causal eGenes.
- Open Access
- 27 April 2026
- Published by Springer Nature in Nature Communications
- Vol. 17
Augmented prediction of multi-species protein–RNA interactions using evolutionary conservation of RNA-binding proteins
RNA-binding proteins (RBPs) play critical roles in the regulation of gene expression. Recent studies have begun to detail the RNA recognition mechanisms of diverse RBPs. However, given the array of RBPs studied so far, it is implausible to experimentally profile RBP-binding peaks for hundreds of RBPs in multiple non-model organisms. Here, we introduce MuSIC ( Mu lti- S pecies RBP–RNA I nteractions using C onservation), a deep learning-based framework for predicting cross-species RBP–RNA interactions by leveraging label smoothing and evolutionary conservation of RBPs across 11 phylogenetically diverse species ranging from human to yeast. MuSIC outperforms state-of-the-art computational methods, and achieves highly accurate prediction of RBP-binding peaks across species. The prediction confidence is higher in the metazoan species, partially reflecting differences in RBP conservation patterns. Finally, the effects of homologous genetic variants on RBP binding can be computationally quantified across species, followed by experimental validations. The target transcripts with disrupted binding events are enriched in the ubiquitination-associated pathways. To summarize, MuSIC provides a useful computational framework for predicting RBP–RNA interactions cross-species and quantifying the effects of genetic variants on RBP binding, offering insights into the RBP-mediated regulatory mechanisms implicated in human diseases.
- 27 April 2026
- Published by Springer Nature in Nature Communications
- Vol. 17
Open AccessAugmented prediction of multi-species protein–RNA interactions using evolutionary conservation of RNA-binding proteins
RNA-binding proteins (RBPs) play critical roles in the regulation of gene expression. Recent studies have begun to detail the RNA recognition mechanisms of diverse RBPs. However, given the array of RBPs studied so far, it is implausible to experimentally profile RBP-binding peaks for hundreds of RBPs in multiple non-model organisms. Here, we introduce MuSIC ( Mu lti- S pecies RBP–RNA I nteractions using C onservation), a deep learning-based framework for predicting cross-species RBP–RNA interactions by leveraging label smoothing and evolutionary conservation of RBPs across 11 phylogenetically diverse species ranging from human to yeast. MuSIC outperforms state-of-the-art computational methods, and achieves highly accurate prediction of RBP-binding peaks across species. The prediction confidence is higher in the metazoan species, partially reflecting differences in RBP conservation patterns. Finally, the effects of homologous genetic variants on RBP binding can be computationally quantified across species, followed by experimental validations. The target transcripts with disrupted binding events are enriched in the ubiquitination-associated pathways. To summarize, MuSIC provides a useful computational framework for predicting RBP–RNA interactions cross-species and quantifying the effects of genetic variants on RBP binding, offering insights into the RBP-mediated regulatory mechanisms implicated in human diseases.
- Open Access
- 5 March 2026
- Published by Springer Nature in Microbiome
- Vol. 14 (1)
Impacts of host genetics on gut microbiome composition in Alzheimer’s disease
Host-microbiome interactions play essential roles in the development of Alzheimer's disease (AD), yet the host genetic impacts on gut microbial alterations in AD remain poorly understood. Here, we simultaneously profiled host genotype and gut microbiome in 252 Chinese individuals with varying degrees of cognitive disability. Using the latent Dirichlet allocation topic model, we identified the Anaerostipes-enriched enterosignature (ES-Ana) at the microbial subgroup level as significantly negatively associated with cognitive disability, which could be recapitulated in external cohorts. With the whole-genome sequencing data, we performed microbiome genome-wide association studies for the ES-Ana relative abundance. We prioritized 41 lead genetic variants and confirmed that the high ES-Ana relative abundance showed a negative correlation with the polygenic risk score of AD, indicating its protective effect against AD. Furthermore, we identified 174 ES-Ana-associated genes, which are enriched in AD-related biological functions and phenotypes, and exhibite pervasive underexpression in glial cells during brain aging. In summary, our study reveals the complex genetic effects on the gut microbiota in AD, and provides novel evidence for the roles of the gut-brain axis in AD. Video
- 5 March 2026
- Published by Springer Nature in Microbiome
- Vol. 14 (1)
Open AccessImpacts of host genetics on gut microbiome composition in Alzheimer’s disease
Host-microbiome interactions play essential roles in the development of Alzheimer's disease (AD), yet the host genetic impacts on gut microbial alterations in AD remain poorly understood. Here, we simultaneously profiled host genotype and gut microbiome in 252 Chinese individuals with varying degrees of cognitive disability. Using the latent Dirichlet allocation topic model, we identified the Anaerostipes-enriched enterosignature (ES-Ana) at the microbial subgroup level as significantly negatively associated with cognitive disability, which could be recapitulated in external cohorts. With the whole-genome sequencing data, we performed microbiome genome-wide association studies for the ES-Ana relative abundance. We prioritized 41 lead genetic variants and confirmed that the high ES-Ana relative abundance showed a negative correlation with the polygenic risk score of AD, indicating its protective effect against AD. Furthermore, we identified 174 ES-Ana-associated genes, which are enriched in AD-related biological functions and phenotypes, and exhibite pervasive underexpression in glial cells during brain aging. In summary, our study reveals the complex genetic effects on the gut microbiota in AD, and provides novel evidence for the roles of the gut-brain axis in AD. Video
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 45 (8) , 1
EIGNN: An Explainable Imaging-Genetic Neural Network for Robust Alzheimer's Disease Risk Prediction
Accurate risk prediction and early diagnosis are crucial for the early intervention of Alzheimer's disease (AD). Current prediction models usually have limited power in capturing the complex interplays between the heterogeneous inputs or lack biological explainability required for clinical adoption and new diagnosis biomarkers discovery. Inspired by pioneering works on biologically informed network and multi-modal learning, we presented an Explainable Imaging-Genetic Neural Network (EIGNN), integrating genetic and neuroimaging data to generate accurate, robust, and explainable AD risk prediction. The EIGNN model features a biologically-informed architecture, incorporating a multi-GWAS SNP selection strategy, enhanced explainable neural network design, and a modal attention mechanism. Genetic variants were hierarchically mapped to their target genes and biological pathways, and further integrated with neuroimaging features. We demonstrated that the EIGNN model outperformed existing methods and exhibited improved robustness, explainability, and reproducibility. Finally, by applying a novel biologically informed multi-modal feature interaction map, we prioritized a set of AD risk genes and biological pathways, and explored the intricate interactions between the risk genes and brain regions implicated in AD risk.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 45 (8) , 1
EIGNN: An Explainable Imaging-Genetic Neural Network for Robust Alzheimer's Disease Risk Prediction
Accurate risk prediction and early diagnosis are crucial for the early intervention of Alzheimer's disease (AD). Current prediction models usually have limited power in capturing the complex interplays between the heterogeneous inputs or lack biological explainability required for clinical adoption and new diagnosis biomarkers discovery. Inspired by pioneering works on biologically informed network and multi-modal learning, we presented an Explainable Imaging-Genetic Neural Network (EIGNN), integrating genetic and neuroimaging data to generate accurate, robust, and explainable AD risk prediction. The EIGNN model features a biologically-informed architecture, incorporating a multi-GWAS SNP selection strategy, enhanced explainable neural network design, and a modal attention mechanism. Genetic variants were hierarchically mapped to their target genes and biological pathways, and further integrated with neuroimaging features. We demonstrated that the EIGNN model outperformed existing methods and exhibited improved robustness, explainability, and reproducibility. Finally, by applying a novel biologically informed multi-modal feature interaction map, we prioritized a set of AD risk genes and biological pathways, and explored the intricate interactions between the risk genes and brain regions implicated in AD risk.
- 10 December 2025
- Published by Springer Nature in Nature Biomedical Engineering
- Vol. 10 (4) , 732-752
A trispecific antibody engaging T cells with tumour and myeloid cells augments antitumour immunity
Immunologically unresponsive tumours often resist immune checkpoint inhibitors due to the low abundance of tumour-specific T cells and an immunosuppressive microenvironment, despite pronounced infiltration of non-tumour-specific (bystander) T cells. Here we analysed single-cell RNA sequencing data from 300 patients across 17 tumour types, identifying abundant but functionally restrained bystander T cells in multiple malignancies, including ovarian and colorectal cancer. To enhance antitumour immunity in such contexts, we engineered B7H3xCD3xPDL1, a trispecific immunoglobulin-based T cell engager targeting B7H3, CD3 and PDL1, to redirect T cells while mitigating immunosuppression. Functional validation in co-culture systems, patient-derived tumour suspensions and fragments, and humanized mouse models showed T cell activation and tumour killing. Imaging cytometry and single-cell transcriptomics revealed IFNγ-dependent macrophage reprogramming and IL-15 secretion, establishing a feed-forward loop that augments T cell functionality. A machine learning model trained on ex vivo cytotoxicity and transcriptomic data predicted patient responsiveness, supporting data-driven clinical stratification for solid tumour immunotherapy.
- 10 December 2025
- Published by Springer Nature in Nature Biomedical Engineering
- Vol. 10 (4) , 732-752
A trispecific antibody engaging T cells with tumour and myeloid cells augments antitumour immunity
Immunologically unresponsive tumours often resist immune checkpoint inhibitors due to the low abundance of tumour-specific T cells and an immunosuppressive microenvironment, despite pronounced infiltration of non-tumour-specific (bystander) T cells. Here we analysed single-cell RNA sequencing data from 300 patients across 17 tumour types, identifying abundant but functionally restrained bystander T cells in multiple malignancies, including ovarian and colorectal cancer. To enhance antitumour immunity in such contexts, we engineered B7H3xCD3xPDL1, a trispecific immunoglobulin-based T cell engager targeting B7H3, CD3 and PDL1, to redirect T cells while mitigating immunosuppression. Functional validation in co-culture systems, patient-derived tumour suspensions and fragments, and humanized mouse models showed T cell activation and tumour killing. Imaging cytometry and single-cell transcriptomics revealed IFNγ-dependent macrophage reprogramming and IL-15 secretion, establishing a feed-forward loop that augments T cell functionality. A machine learning model trained on ex vivo cytotoxicity and transcriptomic data predicted patient responsiveness, supporting data-driven clinical stratification for solid tumour immunotherapy.
- Open Access
- 24 November 2025
- Published by Springer Nature in Genome Biology
- Vol. 26 (1) , 401
HGMT: a database of human gut microbiota for tumors and immunotherapy response
HGMT is a database designed to analyze, explore, and visualize gut microbiomes from diverse tumor types. We process metagenomic datasets from 18,630 stool samples across 37 tumor types, including 2,207 samples from immunotherapy-treated patients across 12 tumor types. HGMT provides an interactive portal for querying taxonomic and functional profiles, visualizing cross-dataset differential abundance taxa in tumors, and identifying their pan-tumor associations. Our analysis reveals the capability of gut microbiota in diagnosing gastrointestinal tumors and predicting immunotherapy response for non-small cell lung carcinoma. HGMT represents a valuable resource for investigating the roles of gut microbiota in tumors and immunotherapy response.
- 24 November 2025
- Published by Springer Nature in Genome Biology
- Vol. 26 (1) , 401
Open AccessHGMT: a database of human gut microbiota for tumors and immunotherapy response
HGMT is a database designed to analyze, explore, and visualize gut microbiomes from diverse tumor types. We process metagenomic datasets from 18,630 stool samples across 37 tumor types, including 2,207 samples from immunotherapy-treated patients across 12 tumor types. HGMT provides an interactive portal for querying taxonomic and functional profiles, visualizing cross-dataset differential abundance taxa in tumors, and identifying their pan-tumor associations. Our analysis reveals the capability of gut microbiota in diagnosing gastrointestinal tumors and predicting immunotherapy response for non-small cell lung carcinoma. HGMT represents a valuable resource for investigating the roles of gut microbiota in tumors and immunotherapy response.
- Open Access
- 8 August 2025
- Published by Oxford University Press (OUP) in Genomics, Proteomics & Bioinformatics
- Vol. 23 (5)
Regulatory Genomic Circuitry of Brain Age by Integrative Functional Genomic Analyses
Brain age gap (BAG) is a valuable biomarker for evaluating brain healthy status and detecting age-associated cognitive degeneration. However, the genetic architecture of BAG and the underlying mechanisms are poorly understood. Here, we estimated brain age from magnetic resonance imaging with improved accuracy using our proposed adversarial convolution network (ACN), and applied the ACN model to an elderly cohort from the UK Biobank. The genetic heritability of BAG was significantly enriched in regulatory regions and implicated in glial cells. We prioritized a set of BAG-associated genes, and further characterized their expression patterns across brain cell types and regions. Two BAG-associated genes, RUNX2 and KLF3, were found to be associated with epigenetic clock and diverse aging-related biological pathways. Finally, two BAG-associated hub transcription factor genes, KLF3 and SOX10, were identified as regulators of pleiotropic risk genes for diverse brain disorders. Altogether, we improve the estimation of BAG, and identify BAG-associated genes and regulatory networks implicated in brain disorders.
- 8 August 2025
- Published by Oxford University Press (OUP) in Genomics, Proteomics & Bioinformatics
- Vol. 23 (5)
Open AccessRegulatory Genomic Circuitry of Brain Age by Integrative Functional Genomic Analyses
Brain age gap (BAG) is a valuable biomarker for evaluating brain healthy status and detecting age-associated cognitive degeneration. However, the genetic architecture of BAG and the underlying mechanisms are poorly understood. Here, we estimated brain age from magnetic resonance imaging with improved accuracy using our proposed adversarial convolution network (ACN), and applied the ACN model to an elderly cohort from the UK Biobank. The genetic heritability of BAG was significantly enriched in regulatory regions and implicated in glial cells. We prioritized a set of BAG-associated genes, and further characterized their expression patterns across brain cell types and regions. Two BAG-associated genes, RUNX2 and KLF3, were found to be associated with epigenetic clock and diverse aging-related biological pathways. Finally, two BAG-associated hub transcription factor genes, KLF3 and SOX10, were identified as regulators of pleiotropic risk genes for diverse brain disorders. Altogether, we improve the estimation of BAG, and identify BAG-associated genes and regulatory networks implicated in brain disorders.
- Open Access
- 7 March 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-15
Spatiotemporally resolved transcriptomics reveals the cellular dynamics of human retinal development
The morphogenesis and cellular interactions in developing retina are incompletely characterized. The full understanding needs a precise mapping of the gene expression with a single-cell spatial resolution. Here, we present a spatial transcriptomic (ST) resource for the developing human retina at six developmental stages. Combining the spatial and single-cell transcriptomic data enables characterization of the cell-type-specific expression profiles at distinct anatomical regions at each developmental stage, highlighting the spatiotemporal dynamics of cellular composition during retinal development. All the ST spots are catalogued into consensus spatial domains, which are further associated to their specific expression signatures and biological functions associated with neuron and eye development. We prioritize a set of critical regulatory genes for the transitions of spatial domains during retinal development. Differentially expressed genes from different spatial domains are associated with distinct retinal diseases, indicating the biological relevance and clinical significance of the spatially defined gene expression. Finally, we reconstruct the spatial cellular communication networks, and highlight critical ligand-receptor interactions during retinal development. Overall, our study reports the spatiotemporal dynamics of gene expression and cellular profiles during retinal development, and provides a rich resource for the future studies on retinogenesis.
- 7 March 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-15
Open AccessSpatiotemporally resolved transcriptomics reveals the cellular dynamics of human retinal development
The morphogenesis and cellular interactions in developing retina are incompletely characterized. The full understanding needs a precise mapping of the gene expression with a single-cell spatial resolution. Here, we present a spatial transcriptomic (ST) resource for the developing human retina at six developmental stages. Combining the spatial and single-cell transcriptomic data enables characterization of the cell-type-specific expression profiles at distinct anatomical regions at each developmental stage, highlighting the spatiotemporal dynamics of cellular composition during retinal development. All the ST spots are catalogued into consensus spatial domains, which are further associated to their specific expression signatures and biological functions associated with neuron and eye development. We prioritize a set of critical regulatory genes for the transitions of spatial domains during retinal development. Differentially expressed genes from different spatial domains are associated with distinct retinal diseases, indicating the biological relevance and clinical significance of the spatially defined gene expression. Finally, we reconstruct the spatial cellular communication networks, and highlight critical ligand-receptor interactions during retinal development. Overall, our study reports the spatiotemporal dynamics of gene expression and cellular profiles during retinal development, and provides a rich resource for the future studies on retinogenesis.
- Open Access
Deciphering causal relationships between cell type-specific genetic factors and brain imaging-derived phenotypes and disorders
The integration of expression quantitative trait loci (eQTLs) and genome-wide association study (GWAS) findings to identify causal genes aids in elucidating the biological mechanisms and the discovery of potential drug targets underlying complex traits. This can be achieved by Mendelian randomization (MR), but to date, most MR studies investigating the contribution of genes to brain phenotypes have been conducted on heterogeneous brain tissues and not on specific cell types, thus limiting our knowledge at the cellular level. In this study, we employ a MR framework to infer cell type-specific causal relationships between gene expression and brain-associated complex traits, using eQTL data from eight cell types and large-scale GWASs of 123 imaging-derived phenotypes (IDPs) and 26 brain disorders and behaviors (DBs). Our analysis constructs a cell type-specific causal gene atlas for IDPs and DBs, which include 254 and 217 potential causal cell type-specific eQTL target genes (eGenes) for IDPs and DBs, respectively. The identified results exhibit high cell type specificity, with over 90% of gene-IDP and 80% of gene-DB associations being unique to a single cell type. We highlight shared cell type-specific patterns between IDPs and DBs, characterize the putative causal pathways among cell type-specific causal eGenes, DBs and IDPs, and reveal the spatiotemporal expression patterns of these cell type-specific causal eGenes. We also demonstrate that cell type-specific causal eGenes can characterize the associations between IDPs and DBs. In summary, our study provides novel insights into the genetic foundations at the cellular level that influence brain structures, disorders and behaviors, which reveals important implications for therapeutic targets and brain health management.
Open AccessDeciphering causal relationships between cell type-specific genetic factors and brain imaging-derived phenotypes and disorders
The integration of expression quantitative trait loci (eQTLs) and genome-wide association study (GWAS) findings to identify causal genes aids in elucidating the biological mechanisms and the discovery of potential drug targets underlying complex traits. This can be achieved by Mendelian randomization (MR), but to date, most MR studies investigating the contribution of genes to brain phenotypes have been conducted on heterogeneous brain tissues and not on specific cell types, thus limiting our knowledge at the cellular level. In this study, we employ a MR framework to infer cell type-specific causal relationships between gene expression and brain-associated complex traits, using eQTL data from eight cell types and large-scale GWASs of 123 imaging-derived phenotypes (IDPs) and 26 brain disorders and behaviors (DBs). Our analysis constructs a cell type-specific causal gene atlas for IDPs and DBs, which include 254 and 217 potential causal cell type-specific eQTL target genes (eGenes) for IDPs and DBs, respectively. The identified results exhibit high cell type specificity, with over 90% of gene-IDP and 80% of gene-DB associations being unique to a single cell type. We highlight shared cell type-specific patterns between IDPs and DBs, characterize the putative causal pathways among cell type-specific causal eGenes, DBs and IDPs, and reveal the spatiotemporal expression patterns of these cell type-specific causal eGenes. We also demonstrate that cell type-specific causal eGenes can characterize the associations between IDPs and DBs. In summary, our study provides novel insights into the genetic foundations at the cellular level that influence brain structures, disorders and behaviors, which reveals important implications for therapeutic targets and brain health management.
- Open Access
- 27 December 2023
- Published by Public Library of Science (PLoS) in PLOS Genetics
- Vol. 19 (12) , e1011112
An augmented Mendelian randomization approach provides causality of brain imaging features on complex traits in a single biobank-scale dataset
Mendelian randomization (MR) is an effective approach for revealing causal risk factors that underpin complex traits and diseases. While MR has been more widely applied under two-sample settings, it is more promising to be used in one single large cohort given the rise of biobank-scale datasets that simultaneously contain genotype data, brain imaging data, and matched complex traits from the same individual. However, most existing multivariable MR methods have been developed for two-sample setting or a small number of exposures. In this study, we introduce a one-sample multivariable MR method based on partial least squares and Lasso regression (MR-PL). MR-PL is capable of considering the correlation among exposures (e.g., brain imaging features) when the number of exposures is extremely upscaled, while also correcting for winner’s curse bias. We performed extensive and systematic simulations, and demonstrated the robustness and reliability of our method. Comprehensive simulations confirmed that MR-PL can generate more precise causal estimates with lower false positive rates than alternative approaches. Finally, we applied MR-PL to the datasets from UK Biobank to reveal the causal effects of 36 white matter tracts on 180 complex traits, and showed putative white matter tracts that are implicated in smoking, blood vascular function-related traits, and eating behaviors.
- 27 December 2023
- Published by Public Library of Science (PLoS) in PLOS Genetics
- Vol. 19 (12) , e1011112
Open AccessAn augmented Mendelian randomization approach provides causality of brain imaging features on complex traits in a single biobank-scale dataset
Mendelian randomization (MR) is an effective approach for revealing causal risk factors that underpin complex traits and diseases. While MR has been more widely applied under two-sample settings, it is more promising to be used in one single large cohort given the rise of biobank-scale datasets that simultaneously contain genotype data, brain imaging data, and matched complex traits from the same individual. However, most existing multivariable MR methods have been developed for two-sample setting or a small number of exposures. In this study, we introduce a one-sample multivariable MR method based on partial least squares and Lasso regression (MR-PL). MR-PL is capable of considering the correlation among exposures (e.g., brain imaging features) when the number of exposures is extremely upscaled, while also correcting for winner’s curse bias. We performed extensive and systematic simulations, and demonstrated the robustness and reliability of our method. Comprehensive simulations confirmed that MR-PL can generate more precise causal estimates with lower false positive rates than alternative approaches. Finally, we applied MR-PL to the datasets from UK Biobank to reveal the causal effects of 36 white matter tracts on 180 complex traits, and showed putative white matter tracts that are implicated in smoking, blood vascular function-related traits, and eating behaviors.