Could two founders regenerate population-level genetic diversity instead of carrying it with them?
I had a slightly outlandish sci-fi thought:
If you were forced to start a human population from only two or few founders, could you engineer a temporary inheritance mechanism that generates diversity for the first few generations, then gradually switches itself off?
Instead of carrying a large frozen genome bank, the founders would carry something closer to a population initialization algorithm.
I expected the idea to fall apart pretty quickly, so I built a simulation and kept removing the obvious cheats.
The later experiments use:
no donor genome bank
no stored source haplotypes
no source allele-frequency table
no list of source polymorphic positions
procedurally generated founder variation
recombination and linkage
blind demographic inference against the resulting population
The interesting result is that the mechanism can repair a surprising amount of the genetic damage from an extreme two-founder bottleneck.
It does not make the bottleneck disappear. Demographic inference can still detect the scar, which is probably the result I find most reassuring.
The project also includes the failed experiments and the sequence of increasingly restrictive models, because most of the interesting work became figuring out where I was accidentally sneaking information into the system.
Repo and experiments:
https://github.com/initsixdev/founder-population-init
I’m very interested in people trying to break the premise.
Where is the biggest biological cheat still hiding?
How do you like the idea?