Control of crossover number:
Several years ago, it seems like a lifetime, I undertook to model outbreeding depression with a computer program. I had been exposed to the language Basic, and supposed that was the way to go. We had been taught to lay out the logic in a diagram with arrows, squares, diamonds and so forth. I more or less laid it out in my mind and then went to buy one of the little plastic things you used to trace out the symbols.
I searched the shelf but could not find the familiar tool. So I asked the fellow behind the counter. He had no idea what I was talking about. Times had changed. I keep having to remind myself that the past was never constant. Things were changing as fast way back when as they are now.
So I asked around and was told that C++ was the best program for number crunching. The rest would be history some day if we were going to have a history. Never mind. I had to teach myself the new way to lay out the logic. Can’t say it was easy.
So I cudgeled my brains. The big break through was when, yes, I could actually create a virtual population with virtual genes and follow it over time. Then I started fooling around with mutations and fertility. I suppose I made every mistake that was possible, including bad logical assumptions, bad spellings, typos and simply getting lost. At its height the monster was working in more than two dozen dimensions and was about twelve ply. That means there would be a logical circuit containing a logical circuit that contained … you get the picture.
Lucky old Einstein. He only had thirteen dimensions and they were all the same kind of animal. These were pretty much all different. And while relativity is laid out in equations that are the same every time, these were generally probability distributions that had to be run instance by instance and no two runs were alike.
All right. I managed to simplify it. I began to get answers.
“Right,” I said, “Now we shall run a number of populations and see that the smaller population is more fertile.”
“Wrong,” said the computer. “The larger is more fertile.” I heaved and hoed and tried everything I could think of. In the end it was a typo that worked. I hadn’t even thought that was a plausible direction.
But I had evidence that could not be denied, and if there is an effect then there must be a cause.
I should have written down the psychosomatic symptoms I ran through. Couldn’t sleep. Couldn’t wake up. Couldn’t eat. Always hungry. The flesh crawlies. The horrors. Nightmares. Dry heaves. Nosebleeds. I think I burned out a half dozen computers; in those days they weren’t designed to be hammered on the way I was doing.
Year after year the computer said, “You’re wrong you’re wrong you’re wrong you’re wrong just like everybody says.”
In the middle of all that I got the notion that the secret might be in recombination. Usually when a cell divides each daughter cell gets a copy of every chromosome the parent cell has. But when the result is to be an egg or sperm there are two replications without making any more chromosomes so that each daughter egg or sperm gets a half set. Mostly the chromosomes occur in pairs and during meiosis homologous chromosomes twist around each other and sometimes break and rejoin, so that aaaaaaaaaa and bbbbbbbbbb become aaaabbbbbb bbbbaaaaaa so to speak. This is crossing over. The fact that some of the a’s are now on the same chromosome as some of the b’s is called recombination.
So I built in the ability to have crossing over. I could control the rate at which it occurred. It only made matters worse. The more crossing over I called for the greater the advantage of the large population over the smaller one. When I finally got the rough number of crossing over events that occurred it became obvious that my little desktop bit cruncher could not possibly handle it so I dropped it.
But I did get firmly in mind that there is indeed an effect of crossing over rate and fertility. And now there is an article (Simona Rosu et al. Robust Crossover Assurance and Regulated Interhomolog Access Maintain Meiotic Crossover Number SCIENCE vol. 334 no. 6060 December 2, 2011 page 1286) that tells us yes indeed the rate must be important because it is tightly regulated.
There have been 38,521 visitors so far.