Siyuan Wu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Siyuan Wu has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Ulf Schmitz
James Cook University
5 shared publications - Tianhai Tian
Monash University 2
3 shared publications Hettiarachchige Wijewardena
Computational Biomedicine Lab, College of Science and Engineering, James Cook University, Townsville, QLD, Australia
2 shared publicationsDebarka Sengupta
Department of Computer Science and Engineering, Indraprastha Institute of Information Technology-Delhi (IIIT-Delhi), Okhla, Phase III, New Delhi 110020, India
1 shared publication- Heather J. Lee
School of Biomedical Sciences and Pharmacy, The University of Newcastle, Callaghan, Australia
1 shared publication Julia Yun-Hsuan Hung
Computational BioMedicine Lab, College of Science and Engineering, James Cook University, Douglas, QLD 4811, Australia
1 shared publication
Latest Publications
- Open Access
A comprehensive benchmark of transcriptome-wide fusion detection using long-read RNA sequencing
Fusion transcripts contribute to cancer, inherited diseases, developmental disorders, and evolution. Long-read RNA sequencing enables direct sequencing of full-length transcripts, creating new opportunities to detect complex fusion architectures, including previously inaccessible multi-segmented fusion transcripts. However, accurate transcriptome-wide fusion detection remains challenging because existing methods struggle to distinguish genuine fusion events from technical artefacts. Here, we present a comprehensive benchmark of transcriptome-wide fusion detection using simulated datasets and transcriptomes from three cancer cell lines across Oxford Nanopore Technologies (ONT) cDNA, PCR-cDNA, and direct RNA sequencing, Pacific Biosciences (PacBio) Kinnex sequencing, Illumina short-read RNA sequencing, six long-read fusion callers, and multiple analysis strategies. False-positive fusion calls remained the dominant limitation across sequencing platforms and algorithms. Increasing sequencing depth improved recall but also amplified spurious fusion calls, whereas higher read-support thresholds improved precision at the expense of sensitivity. ONT PCR-cDNA sequencing combined with CTAT-LR - Fusion achieved the best overall balance between precision and recall, whereas JAFFAL was the only caller to reliably identify simulated tri-gene fusions. Consensus calling reduced false positives but markedly reduced sensitivity, with only one of 400 simulated fusions detected by all six callers. Breakpoint localisation emerged as a major limitation across all methods. Long-read sequencing consistently recovered more validated fusion transcripts than short-read sequencing, enabled detection of complex tri-gene fusions, and produced more biologically plausible fusion landscapes with fewer promiscuous gene partners. Collectively, our results establish the first comprehensive benchmarking framework for transcriptome-wide fusion detection, using long-read RNA sequencing, and provide practical guidance for selecting sequencing workflows and computational strategies, while identifying key priorities for future algorithm development.
Open AccessA comprehensive benchmark of transcriptome-wide fusion detection using long-read RNA sequencing
Fusion transcripts contribute to cancer, inherited diseases, developmental disorders, and evolution. Long-read RNA sequencing enables direct sequencing of full-length transcripts, creating new opportunities to detect complex fusion architectures, including previously inaccessible multi-segmented fusion transcripts. However, accurate transcriptome-wide fusion detection remains challenging because existing methods struggle to distinguish genuine fusion events from technical artefacts. Here, we present a comprehensive benchmark of transcriptome-wide fusion detection using simulated datasets and transcriptomes from three cancer cell lines across Oxford Nanopore Technologies (ONT) cDNA, PCR-cDNA, and direct RNA sequencing, Pacific Biosciences (PacBio) Kinnex sequencing, Illumina short-read RNA sequencing, six long-read fusion callers, and multiple analysis strategies. False-positive fusion calls remained the dominant limitation across sequencing platforms and algorithms. Increasing sequencing depth improved recall but also amplified spurious fusion calls, whereas higher read-support thresholds improved precision at the expense of sensitivity. ONT PCR-cDNA sequencing combined with CTAT-LR - Fusion achieved the best overall balance between precision and recall, whereas JAFFAL was the only caller to reliably identify simulated tri-gene fusions. Consensus calling reduced false positives but markedly reduced sensitivity, with only one of 400 simulated fusions detected by all six callers. Breakpoint localisation emerged as a major limitation across all methods. Long-read sequencing consistently recovered more validated fusion transcripts than short-read sequencing, enabled detection of complex tri-gene fusions, and produced more biologically plausible fusion landscapes with fewer promiscuous gene partners. Collectively, our results establish the first comprehensive benchmarking framework for transcriptome-wide fusion detection, using long-read RNA sequencing, and provide practical guidance for selecting sequencing workflows and computational strategies, while identifying key priorities for future algorithm development.
- Open Access
ICON: An isoform-aware hierarchical random forest model for cell type classification
Single-cell RNA sequencing (scRNA-seq) has transformed our ability to resolve cellular heterogeneity across complex biological systems. However, conventional short-read scRNA-seq is inherently limited in its inability to capture full-length transcripts. Isoform profiles, arising from alternative splicing, provide a deeper layer of resolution, enabling finer discrimination of cellular subtypes and dynamic states, particularly in heterogenous tissues. Long-read RNA sequencing technologies enable accurate transcript-level profiling and more comprehensive characterisation of isoform diversity. Despite these advances, existing cell type annotation methods remain largely tailored to gene-level data, thereby limiting fidelity and leaving isoform-level information an untapped reservoir of biological insight. Here, we present a hierarchical random forest (HRF) framework, ICON, for isoform-aware cell classification in scRNA-seq data. By jointly modelling gene- and isoform-level expression the framework captures both abundance and useage patterns, enabling classification beyond gene-level resolution. A two-stage strategy first assigns cell identities using highly variable gene and isoform features, followed by targeted reclassification of ambiguous cells based on relative isoform and gene usage, thereby resolving conflicts that arise from transcriptional heterogeneity. Importantly, ICON provides interpretable outputs by identifying key genes and isoforms that drive cell type discrimination, linking classification to underlying regulatory mechanisms. Benchmarking on long-read scRNA-seq datasets demonstrates consistent improvements over conventional gene-based approaches. With increasing adoption of long-read sequencing, our framework provides a robust, interpretable foundation for isoform-aware cell type annotation, improving resolution and insight.
Open AccessICON: An isoform-aware hierarchical random forest model for cell type classification
Single-cell RNA sequencing (scRNA-seq) has transformed our ability to resolve cellular heterogeneity across complex biological systems. However, conventional short-read scRNA-seq is inherently limited in its inability to capture full-length transcripts. Isoform profiles, arising from alternative splicing, provide a deeper layer of resolution, enabling finer discrimination of cellular subtypes and dynamic states, particularly in heterogenous tissues. Long-read RNA sequencing technologies enable accurate transcript-level profiling and more comprehensive characterisation of isoform diversity. Despite these advances, existing cell type annotation methods remain largely tailored to gene-level data, thereby limiting fidelity and leaving isoform-level information an untapped reservoir of biological insight. Here, we present a hierarchical random forest (HRF) framework, ICON, for isoform-aware cell classification in scRNA-seq data. By jointly modelling gene- and isoform-level expression the framework captures both abundance and useage patterns, enabling classification beyond gene-level resolution. A two-stage strategy first assigns cell identities using highly variable gene and isoform features, followed by targeted reclassification of ambiguous cells based on relative isoform and gene usage, thereby resolving conflicts that arise from transcriptional heterogeneity. Importantly, ICON provides interpretable outputs by identifying key genes and isoforms that drive cell type discrimination, linking classification to underlying regulatory mechanisms. Benchmarking on long-read scRNA-seq datasets demonstrates consistent improvements over conventional gene-based approaches. With increasing adoption of long-read sequencing, our framework provides a robust, interpretable foundation for isoform-aware cell type annotation, improving resolution and insight.
- Open Access
- 19 March 2026
- Published by Elsevier in Trends in Genetics
- Vol. 42 (8) , 757-770
Domestication pressures and the rise of livestock spliceopathies
- 19 March 2026
- Published by Elsevier in Trends in Genetics
- Vol. 42 (8) , 757-770
Open AccessDomestication pressures and the rise of livestock spliceopathies
- Open Access
- 7 January 2026
- Published by Oxford University Press (OUP) in Bioinformatics Advances
- Vol. 6 (1)
DeepCE: a deep learning framework for correlation-enhanced gene regulatory network inference in single-cell RNA sequencing data
Motivation: Single-cell RNA sequencing has substantially advanced our understanding of gene expression dynamics and cellular heterogeneity. In recent years, deep learning (DL) has emerged as a promising approach to infer genetic regulation. However, these methods still face challenges in representing complex regulatory mechanisms. Thus, it remains imperative to develop new algorithms to enhance both effectiveness and reliability. Results: We propose DeepCE, a DL framework for correlation-enhanced gene regulatory network (GRN) inference. DeepCE strengthens the extraction of dynamic regulation by integrating bidirectional gated recurrent units with convolutional neural networks (CNNs). Specifically, bidirectional gated recurrent units captures dynamic temporal dependencies, while CNNs focuses on local spatial patterns within single-cell data, enabling the model to uncover complex gene-gene interactions and generate high-quality GRNs. This framework improves the accuracy and robustness of GRN inference by smoothing noisy gene expression data, extracting time-lagged regulatory signals, and filtering out spurious correlations. Experiments conducted on mouse and human datasets demonstrate the strong performance of DeepCE. Performance evaluations show that DeepCE outperforms existing methods, achieving the highest AUROC and AUPR scores. Availability and implementation: Codes for DeepCE are free available in the GitHub https://github.com/sxiaodai/DeepCE.
- 7 January 2026
- Published by Oxford University Press (OUP) in Bioinformatics Advances
- Vol. 6 (1)
Open AccessDeepCE: a deep learning framework for correlation-enhanced gene regulatory network inference in single-cell RNA sequencing data
Motivation: Single-cell RNA sequencing has substantially advanced our understanding of gene expression dynamics and cellular heterogeneity. In recent years, deep learning (DL) has emerged as a promising approach to infer genetic regulation. However, these methods still face challenges in representing complex regulatory mechanisms. Thus, it remains imperative to develop new algorithms to enhance both effectiveness and reliability. Results: We propose DeepCE, a DL framework for correlation-enhanced gene regulatory network (GRN) inference. DeepCE strengthens the extraction of dynamic regulation by integrating bidirectional gated recurrent units with convolutional neural networks (CNNs). Specifically, bidirectional gated recurrent units captures dynamic temporal dependencies, while CNNs focuses on local spatial patterns within single-cell data, enabling the model to uncover complex gene-gene interactions and generate high-quality GRNs. This framework improves the accuracy and robustness of GRN inference by smoothing noisy gene expression data, extracting time-lagged regulatory signals, and filtering out spurious correlations. Experiments conducted on mouse and human datasets demonstrate the strong performance of DeepCE. Performance evaluations show that DeepCE outperforms existing methods, achieving the highest AUROC and AUPR scores. Availability and implementation: Codes for DeepCE are free available in the GitHub https://github.com/sxiaodai/DeepCE.
- Open Access
ScIsoX: A Multidimensional Framework for Measuring Isoform-Level Transcriptomic Complexity in Single Cells
Single-cell isoform sequencing enables high-resolution characterisation of transcript isoform expression, yet analytical frameworks to systematically measure transcriptomic complexity are lacking. Here, we introduce ScIsoX , a computational framework that integrates a novel hierarchical data structure, a suite of complexity metrics, and dedicated visualisation tools for isoform-level analysis. ScIsoX supports systematic exploration of global and cell-type-specific isoform expression patterns arising from alternative splicing, revealing multidimensional complexity signatures across diverse datasets - insights often missed by conventional gene-level approaches.
Open AccessScIsoX: A Multidimensional Framework for Measuring Isoform-Level Transcriptomic Complexity in Single Cells
Single-cell isoform sequencing enables high-resolution characterisation of transcript isoform expression, yet analytical frameworks to systematically measure transcriptomic complexity are lacking. Here, we introduce ScIsoX , a computational framework that integrates a novel hierarchical data structure, a suite of complexity metrics, and dedicated visualisation tools for isoform-level analysis. ScIsoX supports systematic exploration of global and cell-type-specific isoform expression patterns arising from alternative splicing, revealing multidimensional complexity signatures across diverse datasets - insights often missed by conventional gene-level approaches.
- Open Access
- 1 January 2025
- Published by American Association for the Advancement of Science (AAAS) in Computational and Structural Biotechnology Journal
- Vol. 27, 4952-4962
Advancing automated cell type annotation with large language models and single-cell isoform sequencing
Accurate cell type identification is critical for interpreting single-cell transcriptomic data and understanding complex biological systems. In this review, we discuss how natural language processing and large language models can enhance the accuracy and scalability of cell type annotation. We also highlight how emerging single-cell long-read sequencing technologies enable isoform-level transcriptomic profiling, offering higher resolution than conventional gene expression-based methods and providing opportunities to redefine cell types. By integrating the insights of key technical and algorithmic advances across sequencing and computational approaches, we provide a unified overview of recent developments that are reshaping automated cell type annotation and improving the precision of biological interpretation.
- 1 January 2025
- Published by American Association for the Advancement of Science (AAAS) in Computational and Structural Biotechnology Journal
- Vol. 27, 4952-4962
Open AccessAdvancing automated cell type annotation with large language models and single-cell isoform sequencing
Accurate cell type identification is critical for interpreting single-cell transcriptomic data and understanding complex biological systems. In this review, we discuss how natural language processing and large language models can enhance the accuracy and scalability of cell type annotation. We also highlight how emerging single-cell long-read sequencing technologies enable isoform-level transcriptomic profiling, offering higher resolution than conventional gene expression-based methods and providing opportunities to redefine cell types. By integrating the insights of key technical and algorithmic advances across sequencing and computational approaches, we provide a unified overview of recent developments that are reshaping automated cell type annotation and improving the precision of biological interpretation.
- Open Access
- 25 March 2022
- Published by Springer Nature in npj Systems Biology and Applications
- Vol. 8 (1) , 1-9
A robust method for designing multistable systems by embedding bistable subsystems
Although multistability is an important dynamic property of a wide range of complex systems, it is still a challenge to develop mathematical models for realising high order multistability using realistic regulatory mechanisms. To address this issue, we propose a robust method to develop multistable mathematical models by embedding bistable models together. Using the GATA1-GATA2-PU.1 module in hematopoiesis as the test system, we first develop a tristable model based on two bistable models without any high cooperative coefficients, and then modify the tristable model based on experimentally determined mechanisms. The modified model successfully realises four stable steady states and accurately reflects a recent experimental observation showing four transcriptional states. In addition, we develop a stochastic model, and stochastic simulations successfully realise the experimental observations in single cells. These results suggest that the proposed method is a general approach to develop mathematical models for realising multistability and heterogeneity in complex systems.
- 25 March 2022
- Published by Springer Nature in npj Systems Biology and Applications
- Vol. 8 (1) , 1-9
Open AccessA robust method for designing multistable systems by embedding bistable subsystems
Although multistability is an important dynamic property of a wide range of complex systems, it is still a challenge to develop mathematical models for realising high order multistability using realistic regulatory mechanisms. To address this issue, we propose a robust method to develop multistable mathematical models by embedding bistable models together. Using the GATA1-GATA2-PU.1 module in hematopoiesis as the test system, we first develop a tristable model based on two bistable models without any high cooperative coefficients, and then modify the tristable model based on experimentally determined mechanisms. The modified model successfully realises four stable steady states and accurately reflects a recent experimental observation showing four transcriptional states. In addition, we develop a stochastic model, and stochastic simulations successfully realise the experimental observations in single cells. These results suggest that the proposed method is a general approach to develop mathematical models for realising multistability and heterogeneity in complex systems.
- Open Access
A non-linear reverse-engineering method for inferring genetic regulatory networks
Hematopoiesis is a highly complex developmental process that produces various types of blood cells. This process is regulated by different genetic networks that control the proliferation, differentiation, and maturation of hematopoietic stem cells (HSCs). Although substantial progress has been made for understanding hematopoiesis, the detailed regulatory mechanisms for the fate determination of HSCs are still unraveled. In this study, we propose a novel approach to infer the detailed regulatory mechanisms. This work is designed to develop a mathematical framework that is able to realize nonlinear gene expression dynamics accurately. In particular, we intended to investigate the effect of possible protein heterodimers and/or synergistic effect in genetic regulation. This approach includes the Extended Forward Search Algorithm to infer network structure (top-down approach) and a non-linear mathematical model to infer dynamical property (bottom-up approach). Based on the published experimental data, we study two regulatory networks of 11 genes for regulating the erythrocyte differentiation pathway and the neutrophil differentiation pathway. The proposed algorithm is first applied to predict the network topologies among 11 genes and 55 non-linear terms which may be for heterodimers and/or synergistic effect. Then, the unknown model parameters are estimated by fitting simulations to the expression data of two different differentiation pathways. In addition, the edge deletion test is conducted to remove possible insignificant regulations from the inferred networks. Furthermore, the robustness property of the mathematical model is employed as an additional criterion to choose better network reconstruction results. Our simulation results successfully realized experimental data for two different differentiation pathways, which suggests that the proposed approach is an effective method to infer the topological structure and dynamic property of genetic regulations.
Open AccessA non-linear reverse-engineering method for inferring genetic regulatory networks
Hematopoiesis is a highly complex developmental process that produces various types of blood cells. This process is regulated by different genetic networks that control the proliferation, differentiation, and maturation of hematopoietic stem cells (HSCs). Although substantial progress has been made for understanding hematopoiesis, the detailed regulatory mechanisms for the fate determination of HSCs are still unraveled. In this study, we propose a novel approach to infer the detailed regulatory mechanisms. This work is designed to develop a mathematical framework that is able to realize nonlinear gene expression dynamics accurately. In particular, we intended to investigate the effect of possible protein heterodimers and/or synergistic effect in genetic regulation. This approach includes the Extended Forward Search Algorithm to infer network structure (top-down approach) and a non-linear mathematical model to infer dynamical property (bottom-up approach). Based on the published experimental data, we study two regulatory networks of 11 genes for regulating the erythrocyte differentiation pathway and the neutrophil differentiation pathway. The proposed algorithm is first applied to predict the network topologies among 11 genes and 55 non-linear terms which may be for heterodimers and/or synergistic effect. Then, the unknown model parameters are estimated by fitting simulations to the expression data of two different differentiation pathways. In addition, the edge deletion test is conducted to remove possible insignificant regulations from the inferred networks. Furthermore, the robustness property of the mathematical model is employed as an additional criterion to choose better network reconstruction results. Our simulation results successfully realized experimental data for two different differentiation pathways, which suggests that the proposed approach is an effective method to infer the topological structure and dynamic property of genetic regulations.