Draft skeleton. Stage scenarios and anchor quotes are in place; the simulator and code panels are not yet wired.
An individual who starts out AB starts as a single cell, and they grow. They end up as a gonad full of haploid cells. About half of those haploid cells will be A and half will be B. Did they reproduce the same? No. The haploid cells are reproducing differentially. Even though each individual was having the same number of children, we still had differential reproduction — and that's how you had evolution. As long as there is a level of differential reproduction, we get evolutionary change. Even if it's not at the level of reproduction we necessarily think of.
— 202_lec08_06
A — One cell, one genome, no internal competition
Start with a single-cell organism. The whole cell is one selection unit. cov(w, z) at the cell level is everything. There is no "within-cell" selection to speak of.
We are just procaryotes that got delusions of grandeur. The way you get from a procaryote to a eukaryote is by squishing some procaryotes together a whole bunch.
— 145_lec07_07
TODO: single-celled baseline sim. Standard population genetics; no within-cell dynamics.
B — Endosymbiosis: now there are subunits inside
Mitochondria, chloroplasts. Each has its own genome with its own replication. The cell-level cov has to compete with within-cell cov on the organellar genomes. Watch what happens when you increase the within-cell variance.
The first time it was even proposed, we didn't know about other endosymbionts. Her name was Lynn Margulis. Everyone thought she was a crazy person. She was right. She did end up having some other crazy ideas — less right. But about this one, she was right.
— 145_lec07_08
TODO: cell-with-organelles sim. Cell-level cov(W_cell, Z_cell) vs within-cell cov on organelle alleles; show heteroplasmy as the within-cell variance.
C — The cell's policing
How does the cell win? By taking over essential organellar functions. The nucleus stole the mitochondrial genes the mitochondria need to replicate — so mitochondria can't break out and overrun the cell. Specify the test: the more genes the cell has taken from the organelle, the smaller the within-cell selection space.
How many of you saw the original Jurassic Park? The whole concept was that they kept one amino acid the dinosaurs couldn't make, so they could control them — they couldn't breed without it. That's what our cells did to the mitochondria and the chloroplasts. We took some of the genes they need to replicate. They can make 99% of what they need, but the other 1% — the few critical proteins — only the nucleus makes now. Which means they can only replicate as fast as the nucleus permits. Otherwise they'd overrun the cell, like a bacterium would.
— 145_lec09_01
TODO: nucleus-policing-mitochondria sim. Gene-transfer-to-nucleus slider; show heteroplasmy variance collapsing.
D — When the cell loses: cancer initiation
A somatic mutation creates a cell lineage with higher proliferation. Within the body, that lineage outcompetes its neighbors. The cell that broke ranks is winning — and the body that contains it is losing. Lesson 30 picks up this same case at the next level up.
This is not cells competing with cells. This is loci in your genome competing with other loci in your genome — because the loci copy themselves, the loci have differential reproduction, the loci have heritable variation. That's all you need for an evolutionary process. So you have loci within a cell evolving, cells within an organism evolving, organisms within a population evolving — all the way up, all at once. And they will be at cross purposes.
— 202_lec15_04
TODO: cancer-initiation sim. Somatic mutation rate; cell-level Price ratio re-emerging within one organism; .R export. Non-trivial code mod: model the policing dose (immune surveillance) as a slider and find where surveillance fails.