[{"content":"I received my Doctor of Philosophy in Computer Science at the University of California, Irvine with my advisor professor Jing Zhang. My research interests include the modeling and analysis of complex, multi-omic high-throughput sequencing data, particularly single-cell and spatial sequencing, using artificial intelligence architectures, with a focus of multiome integration and biological interpretability. I have taught various courses in computer science, data science, and computational biology as a TA. I am a member of the PsychENCODE and the SCORCH consortium.\n","date":"February 19 2026","externalUrl":null,"permalink":"/","section":"Homepage","summary":"","title":"Homepage","type":"page"},{"content":"While it\u0026rsquo;s already passed the New Year, I am about to depart for a new adventure. Hopefully, it will be a bon voyage.\n","date":"February 19 2026","externalUrl":null,"permalink":"/posts/first-post/","section":"Posts","summary":"","title":"New Year, New Start","type":"posts"},{"content":"","date":"February 19 2026","externalUrl":null,"permalink":"/posts/","section":"Posts","summary":"","title":"Posts","type":"posts"},{"content":"","externalUrl":null,"permalink":"/contact/","section":"Homepage","summary":"","title":"","type":"page"},{"content":"","externalUrl":null,"permalink":"/authors/","section":"Authors","summary":"","title":"Authors","type":"authors"},{"content":"","externalUrl":null,"permalink":"/categories/","section":"Categories","summary":"","title":"Categories","type":"categories"},{"content":" Education # Ph.D. in Computer Science,\u0026nbsp;University of California, Irvine\r📍Irvine, CA\rJun. 2021 - Mar. 2026\rComputational Genomics · Interpretable AI · Multi-omic Integration\rDissertation: From Modality Translation to Regulatory Reasoning: Interpretable AI for Single-Cell Multi-Omic Systems Advisor: Prof. Jing Zhang M.S. in Computational Biology,\u0026nbsp;Carnegie Mellon University\r📍Pittsburgh, PA\rAug. 2018 - May. 2020\rCumulative GPA: 3.85 B.S. in Computer Science with a Specialization of Bioinformatics,\u0026nbsp;University of California, San Diego\r📍La Jolla, CA\rSep. 2015 - Jun. 2018\rCumulative GPA: 3.84 / Major GPA: 3.91 Advisor: Awards: Provost Honor, Cum Laude Research # 🏛️Zhang Lab, University of California, Irvine\r📍Computational Genomics · Single-Cell Multi-Omics\r▸Reconstructing Cellular Genomics from Spatial Transcriptomics\r▾\r🗓️Jan. 2025 – Present\r•\rAligned spatial transcriptomics, epigenomics, and histology via multimodal representation learning for in situ states.\r•\rDesigned robust cell-type mapping under sparsity and partial observability across tissues and protocols.\r•\rTrained cross-modality recovery models to denoise and impute gene expression at higher resolution.\r▸The Single-Cell Opioid Responses in the Context of HIV (SCORCH) Consortium\r▾\r🗓️Jan. 2022 – Mar. 2026\r•\rModeled opioid–HIV response programs from consortium-scale single-cell profiles in neural and immune cells.\r•\rBuilt reproducible QC, batch correction, and harmonization pipelines for cross-site integration.\r•\rStandardized data assets: curated matrices, metadata schemas, and visualization-ready outputs for collaborators.\r•\rCoordinated multi-institution analyses and experimental follow-ups to validate computational hypotheses.\r▸Population-Scale Postmortem Brain Multi-Omics for PTSD and MDD\r▾\r🗓️Jan. 2024 – Jun. 2025\r•\rIntegrated snRNA-seq, snATAC-seq, and spatial Xenium across \u0026gt;2M nuclei in cohorts.\r•\rBuilt cross-cohort harmonization to mitigate protocol, site, and donor-level confounding effects.\r•\rReconstructed cell-type-specific regulatory networks linking stress pathology to gene regulation and chromatin.\r•\rDelivered analysis-ready atlases, annotations, and figures for consortium publication and reuse.\r▸scACT: Accurate Cross-Modality Translation for Single-Cell Multiome Data\r▾\r🗓️Jun. 2023 – Oct. 2024\r•\rDeveloped scACT to translate between paired scRNA-seq and scATAC-seq with high fidelity.\r•\rDesigned modality-bridging embeddings robust to sparsity, library-size shifts, and batch effects.\r•\rBenchmarked against state-of-the-art translators across datasets, cell types, and perturbations.\r•\rEnabled imputation of missing modalities to support downstream regulatory inference and integration.\r▸PsychENCODE2 Consortium: Large-Scale Brain Epigenomics and Regulatory Modeling\r▾\r🗓️Sep. 2022 – Jun. 2024\r•\rIntegrated multi-cohort brain single-cell epigenomic datasets for atlas-scale regulatory analysis.\r•\rBuilt reproducible pipelines for QC, harmonization, and confound-aware cross-study integration.\r•\rLearned robust chromatin representations to stabilize cell-state inference under sparsity and batch shifts.\r•\rProduced cell-type annotations and regulatory programs to support consortium figures and publications.\r▸NIH AIM-AHEAD Consortium: ML Workshops and Biomedical Data Curation\r▾\r🗓️Jun. 2023 – Feb. 2024\r•\rCurated genomic and clinical datasets from public repositories; harmonized formats and metadata.\r•\rDeveloped ML curriculum and delivered hands-on workshops for biomedical researchers.\r•\rCoordinated with partner institutions to deploy AI workflows for translational health studies.\r▸Translator: Transfer Learning for Robust scATAC-seq Representation Learning\r▾\r🗓️Jun. 2021 – Jul. 2022\r•\rIntroduced Translator, a transfer-learning framework pretraining scATAC-seq models on high-quality references.\r•\rLearned reusable chromatin embeddings that improved clustering and cell-type annotation under extreme sparsity.\r•\rPerformed domain adaptation to heterogeneous protocols, reducing batch effects and sequencing-depth shifts.\r•\rBenchmarked across cohorts and tissues; delivered stable peak–gene links and regulatory program discovery\r🏛️Carnegie Mellon University\r📍Biomedical Machine Learning · Multimodal Modeling\r▸Computational Models from Clinical and Microbiome Data\r▾\r🗓️Jul. 2020 – Jun. 2021\r•\rSupervisor: Prof. Ziv Bar-Joseph; Collaborator: Astarte Medical\r•\rBuilt ML models predicting neonatal growth trajectories from clinical and microbiome time series.\r•\rModeled missing and irregular EHR records with Hidden Markov Models and imputation.\r•\rSuggested feeding interventions via model-based counterfactual analysis and cohort stratification\r▸Imaging–Genomics Integration with Cross-Modality Data ▾\r🗓️Nov. 2018 – May. 2020\r•\rSupervisor: Prof. Jian Ma\r•\rLinked histology images and transcriptomes via graph neural networks for joint tissue profiling.\r•\rLearned shared multimodal embeddings using autoencoders for alignment and cross-modality retrieval.\r•\rEnabled spatially aware interpretation via feature attribution over learned tissue representations.\r🏛️University of California, San Diego\r📍Biomedical Databases\r▸Biomedical Data Visualization and Database Standardization\r▾\r🗓️Jun. 2016 – Jun. 2018\r•\rSupervisors: Prof. Xiaoqian Jiang, Prof. Shuang Wang\r•\rMapped clinical variables into OMOP CDM schemas to improve interoperability across databases.\r•\rBuilt visualization tools for cohort exploration, missingness auditing, and temporal trend analysis.\r•\rContributed to OHDSI efforts on scalable, global health data standardization and reuse.\rWork Experience # 🏛️University of California, Irvine\r▸Graduate Student Researcher\r▾\r🗓️Sep. 2021 - Mar. 2026\r•\rDeveloped interpretable ML for single-cell multi-omics: translation, integration, and regulatory reasoning.\r•\rBuilt transfer-learning and generative models for modality prediction under sparsity and batch shift.\r•\rCollaborated in PsychENCODE, SCORCH, and PTSD/MDD consortia to deliver pipelines and atlases.\r•\rCo-authored peer-reviewed publications; released reproducible code, datasets, and documentation for reuse.\r•\rMentored junior researchers on model design, evaluation, and scientific communication.\r▸Teaching Assistant\r▾\r🗓️Sep. 2021 - Mar. 2025\r•\rHead TA for ICS 6D Discrete Mathematics and TA for CS 141 Programming Languages in multiple offerings (247+ students).\r•\rDesigned problem sets, quizzes, and exams; aligned objectives, rubrics, and learning outcomes.\r•\rBuilt Gradescope autograding and Q\u0026amp;A workflows to scale feedback for large enrollments.\r•\rLed sections, review sessions, and office-hour systems across in-person and hybrid delivery.\r🏛️Carnegie Mellon University\r▸Research Associate\r▾\r🗓️Jul. 2020 - Jun. 2021\r•\rBuilt predictive models for neonatal growth using longitudinal clinical and microbiome time series.\r•\rEngineered cohort-aware feature pipelines with missingness modeling and temporal alignment of visits.\r•\rApplied probabilistic sequence models to handle irregular sampling and sparse EHR data.\r•\rEvaluated generalization via site-aware splits, calibration, and uncertainty-aware model selection.\r•\rProduced reproducible training/inference workflows and analysis reports for clinical collaborators.\r▸Teaching Assistant\r▾\r🗓️Jan. 2020 - May. 2020\r•\rLed recitations and office hours; supported the rapid transition to remote/hybrid instruction.\r•\rDeveloped programming assignments, solutions, and rubrics aligned with course learning objectives.\r•\rBuilt autograding and feedback workflows to scale timely, consistent grading for large enrollments.\r•\rProvided targeted support on debugging, scientific computing, and quantitative modeling fundamentals\r🏛️University of California, San Diego ▸Tutor (Undergraduate Teaching Assistant)\r▾\r🗓️Mar. 2016 - Mar. 2018\r•\rLed small-group and one-to-one tutoring in programming, data structures, and data-science fundamentals.\r•\rCoordinated weekly with instructors to design sections, refine assignments, and target common misconceptions.\r•\rBuilt fair, rubric-driven assessments; improved clarity of exam questions and scaled feedback with Gradescope.\r•\rProvided inclusive, scaffolded support via flexible office hours and high-touch debugging guidance.\rSkills # Programming: Python, R, Java, C/C++, SQL, Ocaml, Haskell, Prolog, Matlab, WebDev AI/Machine Learning Tools: PyTorch, Tensorflow, Conda, Huggingface Bioinformatics Tools: Seurat, ArchR, Signac, Pegasus, Squidpy, Cellranger, Samtools, Bowtie, David Teaching: Tools Canvas, BlackBoard, Gradescope, Zybooks Languages: Chinese (Mandarin), English, Japanese Services # Journal Reviewer # BioData Mining BMC Bioinformatics BMC Genomics Discover Computing Discover Genetics and Evolution npj Artificial Intelligence Scientific Reports The Journal of Supercomputing Translational Psychiatry Conference Reviewer # 2024 IEEE International Conference on Bioinformatics and Biomedicine 2024-2025 Conference on Information and Knowledge Management 2024-2026 International Conference on Intelligent Systems for Molecular Biology 2026 ACM SIGKDD Conference on Knowledge Discovery and Data Mining Program Committee Member # 2024-2025 International Conference on Artificial Intelligence for Medicine, Health, and Care Awards # Best Student Paper Award, 2024 Asian Conference on Machine Learning Recipient, NSF Student Travel Award ","externalUrl":null,"permalink":"/cv/","section":"Curriculum Vitae","summary":"","title":"Curriculum Vitae","type":"cv"},{"content":" Journal Publications # Deep learning for psychiatric genomics: from tools to applications\u0026nbsp; 📄\rJunhao Liu, Siwei Xu, Dongbo Sun, Chaoyang Wang, Jing Zhang\rCurrent Opinion in Genetics \u0026amp; Development\r2026\r🔗\rdoi: 10.1016/j.gde.2026.102442\rCentral amygdala single-nucleus atlas reveals chromatin and gene transcription dynamics in human alcohol use disorder\u0026nbsp; 📄\rChe Yu Lee et al.\rNature Communications\r2026\r🔗\rdoi: 10.1038/s41467-026-68351-1\rTranscriptomic and Chromatin Dynamics of the Human PTSD Brain at Single Cell Resolution\u0026nbsp; 📄\rMatthew Girgenti et al.\rEuropean Neuropsychopharmacology 2025\r🔗\rdoi: 10.1016/j.euroneuro.2025.08.508\rCombined single-cell profiling of chromatin–transcriptome and splicing across brain cell types, regions and disease state\u0026nbsp; 📄\rWen Hu et al.\rNature Biotechnology\r2025\r🔗\rdoi: 10.1038/s41587-025-02734-5\rSingle cell transcriptomic and chromatin dynamics of the human PTSD brain\u0026nbsp; 📄\rAhyeon Hwang et al.\rNature\r2025\r🔗\rdoi: 10.1038/s41586-025-09083-y\rSingle-cell genomics and regulatory networks for 388 human brains\u0026nbsp; 📄\rPrashant S Emani et al.\rScience\r2024\r🔗\rdoi: 10.1126/science.adi5199\rscENCORE: leveraging single-cell epigenetic data to predict chromatin conformation using graph embedding\u0026nbsp; 📄\rZiheng Duan et al.\rBriefings in Bioinformatics\r2024\r🔗\rdoi: 10.1093/bib/bbae096\rTranslator: A Transfer Learning Approach to Facilitate Single-Cell ATAC-Seq Data Analysis from Reference Dataset\u0026nbsp; 📄\rSiwei Xu et al.\rJournal of Computational Biology\r2022\r🔗\rdoi: 10.1089/cmb.2021.0596\rInsuLock: A Weakly Supervised Learning Approach for Accurate Insulator Prediction, and Variant Impact Quantification\u0026nbsp; 📄\rShushrruth Sai Srinivasan et al.\rGenes\r2022\r🔗\rdoi: 10.3390/genes13040621\rIntegrating longitudinal clinical and microbiome data to predict growth faltering in preterm infants\u0026nbsp; 📄\rJose Lugo-Martinez et al.\rJournal of Biomedical Informatics\r2022\r🔗\rdoi: 10.1016/j.jbi.2022.104031\rConference Publications # Multimodal Cell Context Instruction Tuning for Conditional DNA Regulatory Sequence Generation with Large Language Models\u0026nbsp; 📄\rJunhao Liu et al.\rIEEE International Conference on Image Processing (ICIP)\r2025\r🔗\rdoi: 10.1109/ICIP55913.2025.11084625\rCytoFlow: A Novel Computational Method to Construct Signal Transduction Networks at Single-Cell Resolution based on Flow Networks\u0026nbsp; 📄\rYi Dai, Ziheng Duan, Siwei Xu, Jing Zhang\rIEEE International Conference on Medical Artificial Intelligence (MedAI)\r2024\r🔗\rdoi: 10.1109/MedAI62885.2024.00034\rUnderstanding Transcriptional Regulatory Redundancy by Learnable Global Subset Perturbations\u0026nbsp; 📄\rJunhao Liu et al.\rAsian Conference on Machine Learning (ACML)\r2024\riHAST: Integrating Hybrid Attention for Super-Resolution in Spatial Transcriptomics\u0026nbsp; 📄\rXi Li, Jing Zhang, Ziheng Duan, Yi Dai, Siwei Xu\rBritish Machine Vision Conference (BMVC)\r2024\rLearnable Subset Perturbations for Understanding Transcriptional Regulatory Redundancy\u0026nbsp; 📄\rJunhao Liu, Siwei Xu, Dylan Riffle, Ziheng Duan, Jing Zhang\rNeurIPS 2024 Workshop on Data-driven and Differentiable Simulations, Surrogates, and Solvers\r2024\rscACT: Accurate Cross-modality Translation via Cycle-consistent Training from Unpaired Single-cell Data\u0026nbsp; 📄\rSiwei Xu, Junhao Liu, Jing Zhang\rACM International Conference on Information and Knowledge Management (CIKM)\r2024\r🔗\rdoi: 10.1145/3627673.3679576\riMIRACLE: An Iterative Multi-View Graph Neural Network to Model Intercellular Gene Regulation From Spatial Transcriptomic Data\u0026nbsp; 📄\rZiheng Duan, Siwei Xu, Cheyu Lee, Dylan Riffle, Jing Zhang\rACM International Conference on Information and Knowledge Management (CIKM)\r2024\r🔗\rdoi: 10.1145/3627673.3679574\rConsortium Publications # The single-cell opioid responses in the context of HIV (SCORCH) consortium\u0026nbsp; 📄\rSCORCH Consortium\rMolecular Psychiatry\r2024\r🔗\rdoi: 10.1038/s41380-024-02620-7\rTranscriptomic sex differences in postmortem brain samples from patients with psychiatric disorders\u0026nbsp; 📄\rPsychENCODE Consortium\rScience Translational Medicine\r2024\r🔗\rdoi: 10.1126/scitranslmed.adh9974\rSingle-cell multi-cohort dissection of the schizophrenia transcriptome\u0026nbsp; 📄\rPsychENCODE Consortium\rScience\r2024\r🔗\rdoi: 10.1126/science.adg5136\rA data-driven single-cell and spatial transcriptomic map of the human prefrontal cortex\u0026nbsp; 📄\rPsychENCODE Consortium\rScience\r2024\r🔗\rdoi: 10.1126/science.adh1938\rMassively parallel characterization of regulatory elements in the developing human cortex\u0026nbsp; 📄\rPsychENCODE Consortium\rScience\r2024\r🔗\rdoi: 10.1126/science.adh0559\rCross-ancestry atlas of gene, isoform, and splicing regulation in the developing human brain\u0026nbsp; 📄\rPsychENCODE Consortium\rScience\r2024\r🔗\rdoi: 10.1126/science.adh0829\rUsing a comprehensive atlas and predictive models to reveal the complexity and evolution of brain-active regulatory elements\u0026nbsp; 📄\rPsychENCODE Consortium\rScience Advances\r2024\r🔗\rdoi: 10.1126/sciadv.adj4452\rEvaluating performance and applications of sample-wise cell deconvolution methods on human brain transcriptomic data\u0026nbsp; 📄\rPsychENCODE Consortium\rScience Advances\r2024\r🔗\rdoi: 10.1126/sciadv.adh2588\r","externalUrl":null,"permalink":"/publications/","section":"Publications","summary":"","title":"Publications","type":"publications"},{"content":"","externalUrl":null,"permalink":"/series/","section":"Series","summary":"","title":"Series","type":"series"},{"content":"","externalUrl":null,"permalink":"/tags/","section":"Tags","summary":"","title":"Tags","type":"tags"}]