{
  "id": 12526683,
  "title": "Joint protein, mRNA and DNA sequence design and optimization with nucleotide-level Potts models",
  "url": "https://urgent.news/2026/10/06/joint-protein-mrna-and-dna-sequence-design-and-optimization-with",
  "topic": "science",
  "section": "Science",
  "published": "2026-10-06T00:00:00.000Z",
  "source": {
    "name": "bioRxiv",
    "slug": "biorxiv",
    "url": "https://www.biorxiv.org/content/10.64898/2026.10.05.756407v1?rss=1"
  },
  "original_language": "en",
  "account": "Researchers have developed NuCaliby, a novel model that can simultaneously design amino acid, mRNA, and DNA sequences for a protein at the nucleotide level. This innovative approach offers greater control over the nucleotide sequence by allowing for codon selection during the design process.\n\nNuCaliby utilizes a structure-conditioned Potts model trained on amino acid data, with synonymous codons factored in during the marginalization step. Like the original Caliby model, NuCaliby maintains high designability at the residue level. However, its energy-based framework enables flexible guidance from sequence objectives during inference, incorporating biological proxies into the coding nucleotide sequence.\n\nWhen guided by organism-specific tRNA adaptation, NuCaliby outperforms previous methods in aligning with host tRNA pools, even beyond synonymous sequence space. This is achieved at the expense of self-consistency in the coding sequence.\n\nNuCaliby also allows for the direct embedding of specified RNA motif sequences into coding sequences, a feat typically requiring a combinatorial explosion of amino acids. In laboratory tests, NuCaliby produced proteins that were both expressible and soluble, matching the performance of state-of-the-art methods. Importantly, it offers a higher degree of control and optimization at the nucleotide level.\n\nOverall, NuCaliby extends protein design capabilities beyond the amino acid level, enabling comprehensive optimization across the central dogma of molecular biology - from DNA to RNA to protein. The model's code and pretrained weights are publicly available, paving the way for further research in this domain.",
  "summary": "Protein design typically generates amino acid sequences first and then translates them into DNA post-hoc limiting control over the nucleotide sequence via codon choice at design time. We introduce NuCaliby, a structure-conditioned model that jointly designs amino acid, mRNA, and DNA sequences of a protein at nucleotide resolution. NuCaliby predicts a Potts model over a nucleotide graph, trained…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}