Oncofusion

start / graph / panel-proxe-public-xenograft-repository

Explainerstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q21-index-fusion-models

PRoXe — the one leukaemia xenograft repository that is actually public, and what that buys

also called Public Repository of Xenografts

PRoXe is a collection of leukaemia and lymphoma mouse models that anyone can request, with the patient details published alongside. Most other collections in this field are private to the laboratory that made them.

The same thing, for a scientist

The Public Repository of Xenografts (PRoXe) is a large, publicly available repository of well-characterised leukaemia and lymphoma patient-derived xenografts that engraft orthotopically, distributed with all de-identified information relevant to the primary specimens and the PDXs derived from them. Its design claim is that adequately powered PDX studies can mimic human randomised trials and generate transcriptional, functional and proteomic biomarkers in both treatment-naive and relapsed/refractory disease. For this beat, PRoXe matters less for what it contains than for the contrast it draws: **every other infant KMT2A::AFF1 model resource found in this beat is held by the laboratory that made it, with no catalogue number and no stated distribution route.**

Go deeper
Obtainability is the variable this beat kept hitting

Across the inventory: the eight Perth infant lines have no repository catalogue number; the four KOCL lines have none; BLIN-3 has none; SCMC-L1 has none; the twelve-specimen infant PDX panel is not given a repository accession in its publication; PPTC models are distributed through the consortium. Only SEM (DSMZ ACC-546), RS4;11 (ATCC CRL-1873, DSMZ ACC-508), MV4-11 (ATCC CRL-9591, DSMZ ACC-102) and ALL-PO (a CCLE/DepMap collection line) are catalogued 2.

That is the mechanism by which the field converged on three lines. The lines everyone uses are the lines you can order. It is not a judgement about which lines best model the disease, and it has been read as one for twenty years.

Facts and figures
resourcePublic Repository of Xenografts (PRoXe)
modalityorthotopically engrafting leukaemia and lymphoma PDX
index fusion contentnot established in this beat
observationthree of the four catalogued lines are the three the field uses; the ten uncatalogued lines are the infant ones
catalogued kmt2a aff1 cell lines
[
  {
    "line": "SEM",
    "accession": "CVCL_0095",
    "catalogue": "DSMZ ACC-546"
  },
  {
    "line": "RS4;11",
    "accession": "CVCL_0093",
    "catalogue": "ATCC CRL-1873; DSMZ ACC-508"
  },
  {
    "line": "MV4-11",
    "accession": "CVCL_0064",
    "catalogue": "ATCC CRL-9591; DSMZ ACC-102"
  },
  {
    "line": "ALL-PO",
    "accession": "CVCL_1069",
    "catalogue": "CCLE/DepMap collection; COSMIC-CLP 910944"
  }
]
uncatalogued infant kmt2a aff1 lines
[
  "PER-490",
  "PER-494",
  "PER-485",
  "PER-785",
  "KOCL-45",
  "KOCL-48",
  "KOCL-58",
  "KOCL-69",
  "BLIN-3",
  "SCMC-L1"
]
How this could bite us back
  • Catalogue availability, not biological fidelity, is the strongest predictor of which lines carry this field's conclusions. That is a selection effect on the entire literature and it is invisible in any individual paper.
  • A repository's public metadata is de-identified by design, so donor age may be binned or withheld — meaning even a public repository may not let an outside investigator confirm that a model is infant-derived.
Still open
  1. Does PRoXe hold any infant KMT2A::AFF1 model? The repository publishes its specimen metadata, so this is answerable from its portal and was not answered here.
  2. Would DSMZ or ATCC accept a deposit of the Perth infant lines? Depositing three infant KMT2A::AFF1 lines into a public catalogue would do more for this field's model problem than most experiments.
Every source behind this entry (2)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    PRoXe is a large, publicly available repository of well-characterised leukaemia and lymphoma patient-derived xenografts that undergo orthotopic engraftment, including all de-identified information relevant to the primary specimens and the PDXs derived from them, established to test whether adequately powered PDX trials in mice can define therapeutic activity across heterogeneous tumours.

    The Public Repository of Xenografts Enables Discovery and Randomized Phase II-like Trials in Mice · Townsend EC, Murakami MA, Christodoulou A, et al. · Cancer Cell 2016
    PMID 27070704DOI
    What this evidence cannot show

    Abstract read; full text not read. The abstract does not state the repository's size, its composition by disease or fusion, or whether it contains any infant KMT2A::AFF1 model — so this entry establishes the resource's existence and design, not its contents. Note that a second PubMed record (PMID 27479034) carries the same title and a different DOI; that record has an empty abstract and is the corrected/duplicate entry. Cite 27070704.

    Evidence that argues against this (1)
    • A repository built from adult-hospital leukaemia specimens will contain almost no infant material whatever its size, because infant KMT2A-rearranged ALL is diagnosed at roughly 1 in 25 paediatric ALL cases and paediatric ALL is itself a small fraction of adult haematology intake. Public availability does not compensate for an absent denominator, and this beat did not establish that PRoXe contains any index-fusion infant model. Rokita JL et al. 2019
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    Of the KMT2A::AFF1 cell lines resolved in this beat, only SEM (DSMZ ACC-546), RS4;11 (ATCC CRL-1873, DSMZ ACC-508), MV4-11 (ATCC CRL-9591, DSMZ ACC-102) and ALL-PO (a Cancer Dependency Map / CCLE collection line, COSMIC-CLP 910944) carry a public repository or collection identifier; the eight Perth infant lines, the four KOCL lines, BLIN-3, SCMC-L1, EU-8, B1, BEL-1, Kid92, ALL-2 and AN4;11 carry none in their Cellosaurus records.

    What this evidence cannot show

    Each record fetched by accession on 2026-09-19 and its DR lines read. Absence of a repository cross-reference in Cellosaurus is not proof that a line cannot be obtained — many lines are distributed by direct request and Cellosaurus does not record that. It does establish that these lines cannot be ordered from a catalogue, which is the operative difference.

    Evidence that argues against this (1)
    • Direct request works. The Cellosaurus records for the Perth lines carry a note that the sampling site came from personal communication with the depositing investigator, which is evidence that the laboratory responds to enquiries about these lines and curates them actively. Bairoch A et al. 2025
Connected to