{
  "abstract": "Protein-ligand scoring functions must capture strongly nonlinear, target-dependent binding physics while remaining generalizable beyond their training data, a balance that purely analytic and purely data-driven models each struggle to achieve on their own. We present PATHNet, a physics-informed framework that learns this unknown functional relationship from molecular-mechanics descriptors using a non-monotonic piecewise-linear mapper with a bounded neural-network correction, trained through a variational, least-action-inspired objective and regulated by a renormalization-group-inspired spectral constraint, with model selection performed through multi-start basin optimization. Across seven DUD-E targets, PATHNet achieved the highest mean active inactive/decoy classification accuracy among the methods tested, outperforming three literature baselines representing distinct architectural families. These results support a physics-grounded, variationally constrained approach to neural-network scoring as a practical route toward more interpretable and transferable protein-ligand interaction models.",
  "abstract_shown": true,
  "affiliations": [
    "Sravathi AI Technology Pvt Ltd"
  ],
  "article_type": "New results",
  "arxiv": null,
  "author_total": 4,
  "authors": [
    {
      "name": "B. G. A. S."
    },
    {
      "name": "A. Singh"
    },
    {
      "name": "S. Kanchan"
    },
    {
      "affs": [
        0
      ],
      "name": "N. M. Kulkarni"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/d8g41bzav5csq.bib",
  "cited_by": 0,
  "doi": "10.64898/2026.10.11.758399",
  "fulltext_url": "https://www.biorxiv.org/content/10.64898/2026.10.11.758399v1.full",
  "id": "d8g41bzav5csq",
  "issns": [],
  "landing_url": "https://www.biorxiv.org/content/10.64898/2026.10.11.758399v1",
  "language": null,
  "licence": "CC-BY-NC-ND",
  "licence_label": "CC BY-NC-ND",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by-nc-nd/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "biorxiv",
      "bioinformatics"
    ],
    [
      "openalex",
      "1703|Computational Drug Discovery Methods"
    ]
  ],
  "openalex": "W7222891858",
  "pdf_url": "https://www.biorxiv.org/content/10.64898/2026.10.11.758399v1.full.pdf",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-11",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/d8g41bzav5csq.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:28:20Z",
      "native_id": "10.64898/2026.10.11.758399",
      "source": "biorxiv",
      "url": "https://www.biorxiv.org/content/10.64898/2026.10.11.758399v1"
    }
  ],
  "subjects": [
    {
      "field": "biology",
      "origin": "biorxiv:bioinformatics",
      "subfield": "bioinformatics"
    }
  ],
  "title": "PATHNet: A Thermodynamic Path-Integral-Inspired Neural Network for Protein-Ligand Interaction Prediction",
  "title_text": "PATHNet: A Thermodynamic Path-Integral-Inspired Neural Network for Protein-Ligand Interaction Prediction",
  "type": "preprint",
  "updated_at": "2026-10-30T12:28:20Z",
  "url": "https://research.draconitic.com/paper/d8g41bzav5csq",
  "venue": "bioRxiv",
  "versions": []
}