Acquire
Obtain inspectable files, metadata and license terms. Record checksums and original publication dates.
Evidence case / Gill et al. / 2022
A molecular shift.
A functional question.
A published-study availability audit. No sample-level reanalysis or new experimental result is claimed.
Case CF-001
What the publication makes inspectable.
2022
First published 8 April.
Version of record 21 April.
In vitro
Human dermal fibroblasts.
Not a human treatment trial.
GEO
Accession GSE165180 reported for methylation and RNA-seq.
Open
Outcome pairing and dataset reuse checks remain unresolved.
Four outcome lenses
Read each result on its own terms.
| Lens | Published observation | Boundary |
|---|---|---|
| Molecular | Substantial changes in selected age-associated molecular measurements. | Different clocks need not agree. |
| Function | Collagen recovery and partial migration improvement. | A migration assay is one functional readout. |
| Identity | Fibroblast identity was reacquired after transient reprogramming. | Identity and function require separate assessment. |
| Safety / durability | No clinical benefit is established by this cultured-cell study. | Long-term human safety is outside this evidence. |
Replication & data access
Do not turn samples into donors.
Figure 1 identifies three donor-derived biological replicates for specified conditions. Figure 3’s migration supplement describes four matched starting-population cases; this does not establish four distinct intervention donors. Technical replicates were averaged. The assay-specific hierarchy and cross-assay sample map must be reconciled before a paired analysis.
The paper links methylation and RNA-seq to GSE165180. Direct repository access encountered a verification wall during this review. No raw-data files or individual migration values were downloaded.
Figure 4 reports an outlier excluded from a displayed mean, with tests reported both ways. A reanalysis should prespecify exclusions and retain sensitivity results.
The reproducibility gate
The next evidence is concrete.
Link
Resolve donors, preparations, batches and outcomes. Do not infer pairability from a shared paper.
Evaluate
Freeze the intended use and analysis choices. Compare with a simpler existing workflow.
Current decision: availability audit only. This is useful evidence for method development. A validated Karam benchmark has not yet been demonstrated.
Source: Gill, Parry, Santos, Okkenhaug, Todd, Hernando-Herraez, Stubbs, Milagre & Reik. eLife 11:e71624. Figure legends and data links. Access checked 23 October 2026. No affiliation or endorsement is implied.