Chen (Benjamin) Hsieh

I'm a bioinformatics PhD candidate at the University of Georgia, in CJ Tsai's lab, defending on 22 October. I study poplar genomes from two angles: what happens to extra gene copies after a genome duplication, and how new genes emerge. Both lead back to one question. How does a tree that cannot move cope with environmental stress?

Lately much of my attention has turned to genomic foundation models , and this fall I'm also consulting part-time with the Computational Biology Consulting Group at IOB. What I keep asking is what these models are actually reading. You can rewrite a gene so the protein and the codon composition are both untouched and only the order changes, and what survives that is smaller than the benchmarks suggest.

Outside research I powerlift, play badminton, and practise yoga. I read more manga than I should, mostly while waiting for Claude Code to finish a task, and Katseye has been on repeat while I work.

Get in touch

Grad school has taught me that good research and good communities both get better when people share ideas, try strange things, and learn from each other. If you want to talk about poplar genomes, codon models, or a problem you are stuck on, I would like to hear from you.

I defend on 22 October 2026. My resume has the compressed version of the work, and the publications page has the full list.

Featured right now

Current research · in progress
What Genomic Foundation Models Actually Read
Genomic language models score well on sequence classification. Hold codon composition exactly fixed and most of that advantage disappears — what is left is small, real, and invisible to composition-based screening. Read more →
Open source · on PyPI
Open-Source Research Tooling
SLURM cost reporting, GPU/HPC job templates, and portable agentic research workflows — the infrastructure from my own research, packaged for reuse. Read more →

Projects

View all

A few things people ask

When do you graduate?

I defend on 22 October 2026; the degree is conferred that December.

What are you working toward next?

Industry computational biology and bioinformatics, mostly the evaluation and interpretation of models over biological sequence data. The consulting page has more on the applied side of that.

Is the foundation-model work published?

Not yet. The manuscript is in revision. Until it clears peer review, the numbers on that page are our internal results, not peer-reviewed findings.

What is CBCG?

The Computational Biology Consulting Group at the UGA Institute of Bioinformatics, led by Dr. Casey Bergman. It helps labs with computational problems they'd otherwise improvise. More on the consulting page .

Can I use your code?

Yes. The research tooling is on GitHub , and some of it is on PyPI. If something is broken or undocumented, let me know and I'll see to it.