{
  "abstract": "Chemical language models (CLMs) are prominent for their effectiveness in exploring chemical space and enabling molecular design and engineering. However, while exploring chemical-linguistic space, CLMs suffer from the semantic gap between natural language and molecular representations. This challenge is primarily due to the inherent modeling differences between molecules and texts: molecules operate unified modeling to learn chemical space, while natural language sequentially models their semantic space. Additionally, the limited availability of high-quality text-to-molecule datasets further exacerbates this challenge. To address the problem, we first verified the information bias in molecular representations from different perspectives. We then developed the Heterogeneous Molecular Encoding (HME) framework, a unified molecular encoder compressing the molecular features from fragment sequence, topology, and conformation with Q-learning. To better model chemical-linguistic space, we further constructed the MCMoD dataset, which contains over one million molecules with various conditions, including properties, fragments, and descriptions. Experimentally, HME promotes CLMs to achieve chemical-linguistic sharing space exploration from two aspects: (1) chemical space exploration with linguistic guidance, where HME achieves significant improvements (+37.8% FCD) for molecular design in multiple constraints, even in zero-shot learning scenarios; (2) linguistic space exploration with molecular guidance, where HME generates textual descriptions with high qualities (+11.6% BLEU) for molecules. These results highlight the precision of HME in handling multi-objective and cross-domain tasks, as well as its remarkable generalization capability on unseen task combinations. HME offers a new perspective on navigating the chemical-linguistic sharing space, advancing the potential of CLMs in both fundamental research and practical applications in chemistry.",
  "abstract_shown": true,
  "affiliations": [
    "Beijing Key Laboratory of Brain-inspired Spiking Large Models, School of Computer Science, Peking University, Beijing, China",
    "Guangdong Provincial Key Laboratory of AI for Science and Computing (AI4SC), School of Artificial Intelligence for Science, Shenzhen Graduate School, Peking University, Shenzhen, China",
    "School of Electronic and Computer Engineering, Shenzhen Graduate School, Peking University, Shenzhen, China",
    "Guangdong Provincial Key Laboratory of AI for Science and Computing (AI4SC), School of Artificial Intelligence for Science, Shenzhen Graduate School, Peking University, Shenzhen, China. yuanli-ece@pku.edu.cn",
    "School of Electronic and Computer Engineering, Shenzhen Graduate School, Peking University, Shenzhen, China. yuanli-ece@pku.edu.cn",
    "Beijing Key Laboratory of Brain-inspired Spiking Large Models, School of Computer Science, Peking University, Beijing, China. yhtian@pku.edu.cn",
    "Guangdong Provincial Key Laboratory of AI for Science and Computing (AI4SC), School of Artificial Intelligence for Science, Shenzhen Graduate School, Peking University, Shenzhen, China. yhtian@pku.edu.cn",
    "School of Electronic and Computer Engineering, Shenzhen Graduate School, Peking University, Shenzhen, China. yhtian@pku.edu.cn"
  ],
  "article_type": null,
  "arxiv": null,
  "author_total": 9,
  "authors": [
    {
      "affs": [
        0,
        1,
        2
      ],
      "name": "Liuzhenghao Lv"
    },
    {
      "affs": [
        1,
        2
      ],
      "name": "Hao Li",
      "orcid": "0000-0001-8710-1290"
    },
    {
      "affs": [
        0,
        1
      ],
      "name": "Yu Wang",
      "orcid": "0000-0003-1606-6795"
    },
    {
      "affs": [
        1,
        2
      ],
      "name": "Zijun Chen",
      "orcid": "0000-0001-8642-1604"
    },
    {
      "affs": [
        1,
        2
      ],
      "name": "Zhiyuan Yan"
    },
    {
      "affs": [
        1,
        2
      ],
      "name": "Zongying Lin",
      "orcid": "0009-0001-2035-1970"
    },
    {
      "affs": [
        1,
        2
      ],
      "name": "Yuan Li",
      "orcid": "0000-0002-6059-1485"
    },
    {
      "affs": [
        3,
        4
      ],
      "name": "Li Yuan",
      "orcid": "0000-0002-2120-5588"
    },
    {
      "affs": [
        5,
        6,
        7
      ],
      "name": "Yonghong Tian",
      "orcid": "0000-0002-8329-0888"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/8n9ywbdy038h1.bib",
  "cited_by": 8,
  "doi": "10.1038/s41467-026-74666-w",
  "fulltext_url": null,
  "id": "8n9ywbdy038h1",
  "issns": [
    "2041-1723"
  ],
  "landing_url": "https://doi.org/10.1038/s41467-026-74666-w",
  "language": "en",
  "licence": "CC-BY-NC-ND",
  "licence_label": "CC BY-NC-ND",
  "licence_source": "journal",
  "licence_url": "https://creativecommons.org/licenses/by-nc-nd/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "openalex",
      "3002|Computational Drug Discovery Methods"
    ]
  ],
  "openalex": "W4405957482",
  "pdf_url": "https://www.nature.com/articles/s41467-026-74666-w_reference.pdf",
  "pmcid": "PMC13482991",
  "pmid": "42431910",
  "published": "2026-07-10",
  "publisher": "Nature Portfolio",
  "ris": "https://research.draconitic.com/paper/8n9ywbdy038h1.ris",
  "sources": [
    {
      "first_seen": "2026-11-07T13:21:00Z",
      "native_id": "W4405957482",
      "source": "openalex",
      "url": "https://doi.org/10.1038/s41467-026-74666-w"
    }
  ],
  "subjects": [
    {
      "field": "chemistry",
      "origin": "openalex:3002|Computational Drug Discovery Methods",
      "subfield": "medicinal"
    }
  ],
  "title": "Navigating chemical-linguistic sharing space with heterogeneous molecular encoding",
  "title_text": "Navigating chemical-linguistic sharing space with heterogeneous molecular encoding",
  "type": "peer-reviewed",
  "updated_at": "2026-11-09T13:21:17Z",
  "url": "https://research.draconitic.com/paper/8n9ywbdy038h1",
  "venue": "Nature Communications",
  "versions": []
}