publications

2026

1. 

   Onconeural antigen NOVA1 dysregulates RNA alternative splicing in breast cancer

   Daniel F. Moakley^*, Ye Wang^*, and Chaolin Zhang

   RNA, 2026

   DOI

2. 

   Deep learning unlocks sequence-divergent synthetic promoters to empower Streptomyces natural product engineering

   Qun Zhou^*, Ye Wang^*, Jixin Dong^*, Xiaomin Zhao, Alan J.X. Guo, Xiaowo Wang, and Yunzi Luo

   Metabolic Engineering, 96:420–433, Apr 2026

   DOI

2025

1. 

   Systematic representation and optimization enable the inverse design of cross-species regulatory sequences in bacteria

   Pengcheng Zhang^*, Qixiu Du^*, Ye Wang^*, Lei Wei, and Xiaowo Wang^†

   Nature Communications, 16(1):1763, Feb 2025

   Abs DOI

   Featured Article — Editors’ Highlights in Biotechnology and Methods, Nature Communications

   Long-tail distribution is a common problem in genetic-elements inverse design and engineering tasks based on generative AI. The combination of massive quantification-screen experiments and exploration procedures guided by generative AI efficiently fills in the sequence-activity landscape and significantly improves the performance of predictive and generative models for inverse design. The strategy of generative AI-guided high-throughput exploration shows potential to overcome data scarcity, a common problem in inverse design and engineering tasks. DeepCROSS offers the potential to go beyond designing bacterial regulatory sequences and be used for mammalian regulatory sequence design, taking a step toward constructing complex circuits using generative AI for the synthetic biology community.

2024

1. 

   Reference-informed prediction of alternative splicing and splicing-altering mutations from sequences

   Chencheng Xu^*, Suying Bao^*, Ye Wang, Wenxing Li, Hao Chen, Yufeng Shen, Tao Jiang^†, and Chaolin Zhang^†

   Genome Research, 34(7):1052–1065, Mar 2024

   Abs DOI

   We introduced a reference-informed computational framework that systematically links sequence context to splice-site usage across human tissues, enabling quantitative prediction of splicing-altering mutations. This platform supports functional interpretation of genetic variants in the context of human disease, providing mechanistic insight into how noncoding and coding variants disrupt RNA processing. By transforming large-scale transcriptomic data into predictive models of splicing regulation, our work facilitates clinically meaningful variant prioritization, improves the interpretation of patient genomic data, and informs the rational design of RNA-targeted therapeutic strategies. This framework bridges fundamental RNA biology with translational medicine, creating a scalable foundation for precision diagnostics and RNA-based therapeutic development.

2. 

   GPro: generative AI-empowered toolkit for promoter design

   Haochen Wang, Qixiu Du, Ye Wang, Hanwen Xu, Zheng Wei, and Xiaowo Wang

   Bioinformatics, 40(3):btae123, Mar 2024

   DOI

2022

1. 

   DeSP: a systematic DNA storage error simulation pipeline

   Lekang Yuan, Zhen Xie, Ye Wang, and Xiaowo Wang

   BMC Bioinformatics, 23(1):185, Mar 2022

   DOI

2020

1. 

   Synthetic promoter design in Escherichia coli based on a deep generative network

   Ye Wang^*, Haochen Wang^*, Lei Wei, Sichao Li, Lingxue Liu, and Xiaowo Wang^†

   Nucleic Acids Research, 48(12):6403–6412, Jan 2020

   Abs DOI

   Cover Paper & Breakthrough Article, Nucleic Acids Research

   My research during PhD training focused on developing AI-based generative algorithms and optimization methods for de novo promoter sequence design. Understanding the complex encoding syntax in the genome and using it to build synthetic circuits will accelerate synthetic biology parts discovery and hopefully bring new drug candidates to the clinic. This approach achieved high success rates in in vivo experimental validations, offering valuable insights into synthetic element design and highlighting the potential of deep learning to explore the vast sequence space of synthetic biology.