{
  "id": 3769407,
  "title": "reactifpTM: an accessible reimplementation of actifpTM",
  "url": "https://urgent.news/2026/08/27/reactifptm-an-accessible-reimplementation-of-actifptm",
  "topic": "science",
  "section": "Science",
  "published": "2026-08-27T00:00:00.000Z",
  "source": {
    "name": "bioRxiv",
    "slug": "biorxiv",
    "url": "https://www.biorxiv.org/content/10.64898/2026.08.24.746624v1?rss=1"
  },
  "original_language": "en",
  "account": "Motivation: The actifpTM score, a modified version of the ipTM score, is a commonly used measure for assessing interface quality. However, it relies on predicted aligned error (PAE) with probabilities, which is generated during a ColabFold run but not output by the package or other model prediction software. This limitation has hindered the adoption of actifpTM for software like AlphaFold 2 or AlphaFold 3, as it cannot generate actifpTM scores for their results. To address this issue, reactifpTM has been developed as a standalone tool. It utilizes standard output files from model prediction software, such as a model and corresponding PAE, to perform an actifpTM-like calculation.\n\nResults: By employing the same underlying principles as actifpTM, reactifpTM uses standard output files from model prediction software to carry out the calculation. When applied to a dataset of 1079 known interfaces in the PDB, strong correlations were observed between the actifpTM and reactifpTM scores.\n\nAvailability and implementation: The reactifpTM tool is coded in Python, making it accessible and easy to use. All scripts and associated documentation are available from the GitHub repository at https://github.com/hlasimpk/reactifptm or through the Python Package Index (PyPI) at https://pypi.org/project/reactifptm.",
  "summary": "Motivation: The actual interface pTM score (actifpTM) is a modified version of the ipTM score that limits the calculation to only those residues at the interface. Whilst actifpTM provides an effective interface quality score, a limiting factor is that it makes use of the predicted aligned error (PAE) with probabilities, information that is generated during a ColabFold run, but not output by the…",
  "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."
}