Senior Deep Learning Scientist, Illumina AI Lab
Kuan-Hao Chao
I build machine learning for genomics — sequence-to-function models, genome annotation, and DNA language models.
I am a Senior Deep Learning Scientist at the Illumina AI Lab. I earned my Ph.D. in Computer Science from the Center for Computational Biology, Johns Hopkins University (August 2025), advised by Steven Salzberg and Mihaela Pertea. My research focuses on AI for genomics — sequence-to-function modeling, genome annotation, and DNA language models. I hold a B.S. in Electrical Engineering from National Taiwan University and exchanged in my final year at the Australian National University.
What I work on
Research
Selected work
Publications
“Build what you need, use what you build.”
Background
Experience & education
Experience
- Aug 2025 – Present Sr. Deep Learning Scientist & Engineer Illumina , AI Lab
- May 2024 – Aug 2024 Genomics ML Research Intern Calico Life Sciences , Kelley Lab
- Jul 2020 – Jan 2021 Research Assistant Academia Sinica , Institute of Information Science
- Jul 2019 – Jun 2020 Research Student Australian National University , Research School of Biology
- Aug 2018 – Jul 2019 Research Student National Taiwan University , Centers of Genomic and Precision Medicine
Education
- Sep 2021 – Aug 2025 Ph.D., Computer Science Johns Hopkins University
- Sep 2021 – May 2023 M.S.E., Computer Science Johns Hopkins University
- Sep 2016 – Jan 2021 B.S., Electrical Engineering National Taiwan University
Writing
Posts
lifton2: a first-party, refinement-first genome-annotation lift-over engine
lifton2 is a research re-implementation of the LiftOn lift-over engine that de-vendors the Liftoff DNA-lift into first-party code, then refines it with protein evidence — improving the genes it transfers, emitting standards-clean GFF3, and finishing whole genomes other tools crash on.
Read postShorkie: learning yeast regulatory code from related fungi
Shorkie tests how fungal pretraining, evolutionary scale, and fine-tuning on yeast regulatory assays help a compact DNA model read expression and variant effects.
Read postUpdates
Recent news
I give an invited talk in the Machine Learning in Genomics session at ProbGen 2026 at UC Berkeley. The talk is titled “Predicting dynamic expression patterns in budding yeast with a fungal DNA language model.”
The OpenSpliceAI version of record is published in eLife. The GitHub repository and documentation are available.
Our Shorkie preprint describing a fungal DNA language model for predicting RNA-seq coverage from DNA sequence in S. cerevisiae is available on bioRxiv.
My public dissertation talk, “Decoding the Language of Genomes: Bridging Sequences and Function through Deep Learning,” is scheduled for August 25, 2025.
I defend my Ph.D. dissertation at Johns Hopkins University on August 25, 2025, advised by Steven Salzberg and Mihaela Pertea.
Let's work together
I'm always glad to talk about computational genomics, machine learning, or mentoring research projects. Reach out anytime.