Open Channel · Transmission Ready

Send a
signal.

Every collaboration starts with a message. Whether it's support, science, or a shared conviction that biology is one problem — we read everything.

Channels
Technical Support
Model errors, API questions, inference issues, environment setup, Colab troubleshooting.
support@terminalbio.com
Collaboration
Research partnerships, academic collaboration, joint model development, dataset sharing.
support@terminalbio.com
Science Queries
Architecture questions, benchmark methodology, model limitations, citation and reproducibility.
support@terminalbio.com
General Inquiry
Everything else — media, feedback, ideas, or just something you needed to say.
support@terminalbio.com
Typical Response Time
Immediate ≤ 24 h 72 h+
Transmission Status
Endpointsupport@terminalbio.com
ProtocolSMTP · TLS 1.3
Status● OPEN
LabIndependent · MIT Licensed
Message Type
Institution, lab, or self-described
Encoding payload…
Signal received.

Your message is en route to support@terminalbio.com. We'll respond within 24 hours.


Common Questions
Can I run BindFM on a free Colab GPU? +
Yes. BindFM v0.1 is designed to run on a T4 GPU in under 40 minutes. The Colab notebook is linked from the GitHub repo. If you hit OOM errors, reduce the batch size or switch to the lightweight checkpoint.
Which binding modalities does BindFM support? +
All five: Protein–Small Molecule, Protein–Protein, Protein–Nucleic Acid, Nucleic–Small Molecule, and Nucleic–Nucleic. Every partner pair is encoded through the same shared EGNN weights.
Can XUZU design aptamers without SELEX data? +
Yes — that's the point. XUZU is a zero-shot aptamer design model. You supply a target protein sequence or structure, and it generates DNA or RNA aptamer candidates ranked by predicted Kd, no experimental SELEX cycles required.
What license covers terminalBio models? +
All models — BindFM, XUZU, RNJ — are released under the MIT license. You can use them freely for research, commercial applications, or as a base for further development with no restrictions.
How do I cite terminalBio models in a paper? +
Citation instructions are in each model's GitHub README. Email us at support@terminalbio.com if you're preparing a preprint and want to coordinate — we're happy to cross-reference and support reproducibility efforts.
Are you open to academic partnerships? +
Absolutely. Use the Collaboration type in the contact form above and describe your lab, the specific research direction, and what a partnership would look like. We're particularly interested in wet-lab validation collaborations.