Regeneration
vs scarring
Why can a mouse regrow a fingertip but scar over a skin wound? I searched public data from two mouse regeneration systems for genes that show up in both. Scroll in, drag the models, hover any part.
independent computational projectSame finger, different cut
In mice, an amputation through the end of the terminal bone (P3) regrows a tip. An amputation one bone back (P2) does not: the wound closes over. These two cases are my digit datasets.
Scar or regenerate?
Skin wounds usually scar. Inhibiting YAP pushes them to regenerate instead. A scar's collagen typically lines up in parallel, while regenerated skin looks more like normal skin, basket-weave and hair follicles included.
This skin dataset has only 3 pooled libraries per group, so I report skin results as descriptive rankings, not significance tests.
Four genes in both
I looked for genes that rank highly in the skin wounds and are specific to digit regeneration. Four survive: Dusp4, Nfatc1, Saa3 and Sfrp4. Nfatc1 and Sfrp4 already have independently published roles in wound regeneration.
A bigger net: 108 genes
Starting from pathways instead, twelve enriched pathways are shared, including ECM organization and PI3K-Akt signaling, with 108 genes underneath. Their STRING network has 457 edges and contains one gap junction gene, Gja1 (connexin 43). That flag is why I built the bioelectric simulation next.
The node layout here is illustrative.
Pre-registered vs exploratory
I kept the two tiers apart and did not weight them equally. The pre-registered 4-gene classifier did not clear a permutation test (p = 0.078). The exploratory 155-gene classifier did (p = 0.020), but it is unvalidated.
What this does not show
Sample sizes are small throughout, the skin data is limited to 3 pooled libraries per group, and every finding is correlational. I make no causal claims. The 3D models, network layout and distributions are schematic.