Sounds theoretically possible. I mean it definitely happens at conception, and methylation happens at specific sites where it should, and not where it shouldn’t. So there should be a mechanism.

Even if it’s trickier than that, there are probably lots of ways to selectively restore the epigenetic information in specific patterns or specific sites that might help chip away at disease.

Somewhat related link from quick googling:

https://www.sciencedirect.com/science/article/pii/S002192582...