{
  "abstract": "Motivation\n\nAccurate Drug-Target Interaction (DTI) prediction is critical for drug discovery. While deep learning advances this field, the fundamental lack of verified negative interactions inevitably causes biased risk estimation. Although formulating DTI prediction as a Positive-Unlabeled (PU) learning problem is theoretically sound, directly applying PU learning to deep models suffers from early-stage optimization instability and decision boundary ambiguity.Results\n\nWe propose ProDTI, a prior-guided PU learning framework integrating domain knowledge into the PU objective. We construct a hybrid prior combining explicit biochemical similarities with implicit semantic representations from large-scale pre-trained models. This prior drives two synergistic mechanisms: soft-label distillation to stabilize the optimization trajectory, and hard-anchor regularization to explicitly penalize reliable negatives and sharpen decision boundaries. Evaluated on five benchmarks, ProDTI consistently outperforms state-of-the-art baselines, exhibiting superior robustness in challenging cold-start scenarios and providing a reliable tool for computational drug discovery.Availability and implementation\n\nSource code and data are available at https://github.com/WangYu-AI4SCI/ProDTI.",
  "abstract_shown": true,
  "affiliations": [
    "School of Artificial Intelligence, China Pharmaceutical University, Nanjing, 211198, China.",
    "School of Computer Science, Nanjing University of Posts and Telecommunications, Nanjing, 210023, China.",
    "Institute of Innovative Drug, China Pharmaceutical University, Nanjing, 211198, China.",
    "School of Pharmacy, China Pharmaceutical University, Nanjing, 211198, China."
  ],
  "article_type": "research article",
  "arxiv": null,
  "author_total": 5,
  "authors": [
    {
      "affs": [
        0
      ],
      "name": "Yu Wang",
      "orcid": "0000-0003-1606-6795"
    },
    {
      "affs": [
        1
      ],
      "name": "Haoxiang Zhang"
    },
    {
      "affs": [
        1
      ],
      "name": "Ziye Zhu"
    },
    {
      "affs": [
        2,
        3
      ],
      "name": "Qiang Wei"
    },
    {
      "affs": [
        1
      ],
      "name": "Yun Li"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/dcavzkv3pzq49.bib",
  "cited_by": 0,
  "doi": "10.1093/bioinformatics/btag712",
  "fulltext_url": "https://europepmc.org/articles/PMC13676319",
  "id": "dcavzkv3pzq49",
  "issns": [
    "1367-4803",
    "1367-4811"
  ],
  "landing_url": "https://europepmc.org/article/MED/42842427",
  "language": "eng",
  "licence": "CC-BY",
  "licence_label": "CC BY",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "openalex",
      "1703|Computational Drug Discovery Methods"
    ]
  ],
  "openalex": "W7220838013",
  "pdf_url": "https://europepmc.org/articles/PMC13676319?pdf=render",
  "pmcid": "PMC13676319",
  "pmid": "42842427",
  "published": "2026-10-01",
  "publisher": "Oxford University Press (OUP)",
  "ris": "https://research.draconitic.com/paper/dcavzkv3pzq49.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:18:10Z",
      "native_id": "MED:42842427",
      "source": "europepmc",
      "url": "https://europepmc.org/article/MED/42842427"
    },
    {
      "first_seen": "2026-10-30T12:19:38Z",
      "native_id": "10.1093/bioinformatics/btag712",
      "source": "crossref",
      "url": "https://academic.oup.com/bioinformatics/article/doi/10.1093/bioinformatics/btag712/8875880"
    }
  ],
  "subjects": [
    {
      "field": "biology",
      "origin": "issn:1367-4803",
      "subfield": "bioinformatics"
    }
  ],
  "title": "ProDTI: prior-guided positive-unlabeled learning for robust drug-target interaction prediction.",
  "title_text": "ProDTI: prior-guided positive-unlabeled learning for robust drug-target interaction prediction.",
  "type": "peer-reviewed",
  "updated_at": "2026-10-30T12:19:38Z",
  "url": "https://research.draconitic.com/paper/dcavzkv3pzq49",
  "venue": "Bioinformatics (Oxford, England)",
  "versions": []
}