{
  "abstract": "We develop an All-atom Structure Tokenizer that compresses biomolecular geometry into discrete structural tokens and achieves high-accuracy all-atom reconstruction. A hierarchical geometric encoder and finite scalar quantization map chemical units to a shared vocabulary of 8,192 tokens, while a joint coordinate de-tokenizer reconstructs backbone, side-chain, nucleic-acid, and ligand heavy atoms. The tokenizer achieves all-atom lDDT scores of 0.832 and 0.835 on CASP14 and CASP15, respectively. Building on this tokenizer, we train AtomLM-9b, an autoregressive language model based on Qwen3.5-9B that predicts structural tokens from molecular identities and connectivity and decodes them into complete atomic structures. We demonstrate, for the first time, protein structure prediction performance comparable to AlphaFold 3 with an all-atom-tokenizer-based structural language model: on the FoldBench protein-protein task, AtomLM-9b achieves a success rate of 75.01%, compared with the published AlphaFold 3 result of 72.93%. These results establish discrete all-atom tokenization as an effective interface between high-fidelity atomic reconstruction and language-model-based structure prediction.",
  "abstract_shown": true,
  "affiliations": [
    "Scinetics Company"
  ],
  "article_type": "New results",
  "arxiv": null,
  "author_total": 5,
  "authors": [
    {
      "name": "J. Yan"
    },
    {
      "name": "D. Xie"
    },
    {
      "name": "Z. Wang"
    },
    {
      "name": "M. Xu"
    },
    {
      "affs": [
        0
      ],
      "name": "Z. Zhang"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/e2xx30th56rgc.bib",
  "cited_by": 0,
  "doi": "10.64898/2026.11.03.763913",
  "fulltext_url": "https://www.biorxiv.org/content/10.64898/2026.11.03.763913v1.full",
  "id": "e2xx30th56rgc",
  "issns": [],
  "landing_url": "https://www.biorxiv.org/content/10.64898/2026.11.03.763913v1",
  "language": null,
  "licence": "CC-BY",
  "licence_label": "CC BY",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "biorxiv",
      "bioinformatics"
    ]
  ],
  "openalex": null,
  "pdf_url": "https://www.biorxiv.org/content/10.64898/2026.11.03.763913v1.full.pdf",
  "pmcid": null,
  "pmid": null,
  "published": "2026-11-03",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/e2xx30th56rgc.ris",
  "sources": [
    {
      "first_seen": "2026-11-05T00:21:58Z",
      "native_id": "10.64898/2026.11.03.763913",
      "source": "biorxiv",
      "url": "https://www.biorxiv.org/content/10.64898/2026.11.03.763913v1"
    }
  ],
  "subjects": [
    {
      "field": "biology",
      "origin": "biorxiv:bioinformatics",
      "subfield": "bioinformatics"
    }
  ],
  "title": "All-Atom Structure Tokenizer: AtomLM-9b for Biomolecular Structure Prediction",
  "title_text": "All-Atom Structure Tokenizer: AtomLM-9b for Biomolecular Structure Prediction",
  "type": "preprint",
  "updated_at": "2026-11-06T00:21:42Z",
  "url": "https://research.draconitic.com/paper/e2xx30th56rgc",
  "venue": "bioRxiv",
  "versions": []
}