{
  "created_utc": "2026-09-19T06:17:10.573192+00:00",
  "queries": [
    "\"Haller\" \"epigenetic\" \"overlap\" training",
    "\"Independent Avian Epigenetic Clocks\" correction validation",
    "\"epigenetic age\" \"brood\" \"training\" leakage",
    "\"epigenetic\" \"training samples\" \"treatment\" \"bias\" clock"
  ],
  "nearest_prior_work": [
    {
      "title": "Haller et al. 2025",
      "url": "https://doi.org/10.1111/1755-0998.14128",
      "finding": "The original notebook openly contains the full-model prediction workflow with 33 overlapping age-14 samples. The article does not quantify prediction differences by membership or provide a brood-excluded treatment assessment. The overlap is derived from data/code, not concealed provenance."
    },
    {
      "title": "Methods for analysing wildlife DNA methylation data",
      "url": "https://doi.org/10.1093/conphys/coaf091",
      "finding": "Explicitly discusses training/test reuse during age-acceleration residual calibration and suggests cross-fitting. No Haller/great-tit brood reuse case or corresponding numerical results found in full text."
    },
    {
      "title": "Designing epigenetic clocks for wildlife research",
      "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC12415925/",
      "finding": "General guidance on independent validation, age bias and wildlife clock design. This principle is not our contribution."
    },
    {
      "title": "The BEAC, an epigenetic clock for birds, August 2026 preprint",
      "url": "https://doi.org/10.64898/2026.08.14.744821",
      "finding": "Independent king penguin test set, technical replicates and cross-species validation. Direct credit as stronger avian validation design; not the specific great-tit analysis."
    }
  ],
  "source_citation_screen": "Europe PMC returned two distinct newer papers plus one preprint version, on fish/shark methylation. No direct reuse critique identified. Citation list retained separately.",
  "conclusion": "Provisional novelty only for an empirically quantified great-tit case and practical reanalysis, not for cross-validation, leakage, cross-fitting or the generic risk of reusing training samples. A second search after quantitative validation is required.",
  "limits": "Search indexing incomplete; no author confirmation or claim that every prior document has been found."
}
