Frequently Asked Questions

Common Questions

What molecules can be designed using AutoNA?

mRNA is supported today—end‑to‑end design plus model‑backed scoring. Additional nucleic‑acid modalities are on our roadmap.

What is the maximum length for an mRNA design?

Your input CDS can be up to 2,000 nucleotides on the AutoNA server. Total transcript length may vary slightly with your chosen UTRs. For best performance, we recommend keeping coding regions ≤ 1500 nt.

What are the components of the mRNA that can be designed?

AutoNA (powered by mRNAutilus) can co‑design the 5' UTR, CDS, and 3' UTR. The 5' cap and poly-(A) tail are not currently designed by mRNAutilus.

What are the differences between each tier?

Your plan sets your limits:

Academic Free: Single-objective, 3 jobs per day
Academic Base: Single-objective 10 jobs per day
Academic Plus: Multi-objective, 10 jobs per day
Commercial Base: Contact us
Commercial Plus: Contact us

Reach out to us at founders@atombioworks.com to increase your cap!

How do I define inputs for mRNA design jobs?

A job definition includes: 1. CDS — begins with a start codon (ATG), ends with a stop codon (TAA/TAG/TGA), and has a length divisible by 3. 2. Optimization strategies — up to three properties (by tier): predicted cellular half‑life, translation efficiency, protein abundance. 3. Number of sequences — how many candidates to generate. 4. Codon optimization scope — optimize a short window (e.g., first 30 codons) or the entire CDS. 5. UTR lengths — choose 5' and 3' UTR lengths. Note: generated 5' UTRs include a shortened Kozak motif (GCCACC); for example, if you request a 50‑nt 5' UTR, the generated length will be ~56 nt.

How do I interpret the scores and quality of a generated sequence?

All predictions were derived from models trained on data collected in human cells. We define the output scores as the following:

Half-life: Log-10 predicted half-life (in hours)
Protein Abundance: Log-10 predicted Protein Quantity per mRNA per hour
Translation efficiency: Predicted normalized ribosome sequencing output

Each job includes a baseline reference sequence for use as a benchmark. It is constructed as the input CDS with human α-globin UTRs.

How can I rerun a job?

If your job fails, you have an option to rerun your job. Click the button at the bottom of the job output page and your job will be re-queued.

How should I cite AutoNA/mRNAutilus?

Please cite the following:

@misc{patel2026mrnautilusmultiobjectiveguideddiscretegeneration,
      title={mRNAutilus: Multi-Objective-Guided Discrete Generation of mRNA with Optimized Therapeutic Properties}, 
      author={Sawan Patel and Sophia Tang and Yesol Kim and Yinuo Zhang and Divya Srijay and Ping-Jung Lin and Shambhavi Shubham and Fengmei Pi and Cedric Wu and Sherwood Yao and Pranam Chatterjee},
      year={2026},
      eprint={2605.31296},
      archivePrefix={arXiv},
      primaryClass={q-bio.BM},
      url={https://arxiv.org/abs/2605.31296}, 
}

Who should I contact for inquiries and general feedback?

All questions can be directed to founders@atombioworks.com. You can also find us at https://www.atombioworks.com/

How can I access the code and weights for mRNAutilus?

Reach out to founders@atombioworks.com for questions regarding mRNAutilus code and weights.