For a breeding program for selective traits you need 1) As much info you can gather about the traits of interest. This tells you what pit falls to look for.
2) You also want as much back history from your breeding subjects as you can get.... What looks genetic isn't always the case: Just because a person has a trait does not mean he can actually pass the trait on (somatic mutation, maternal epigenetic effects, etc. etc. etc.) 3) In any breeding program you want to start with the most far apart set of traits you can find, so you can more correctly track the traits you want (molecularly speaking). Sure breeding bright animals to bright animals will give you a couple of generations of brighter animals but then your population may start suffering - homozygosity in a population isn't a good thing (look up hybrid vigor on wiki). More far apart traits are more likely going to give hybrid vigor (hopefully) but also a clear picture of what trait is from the maternal source and what is from the paternal source thus allowing you to develop a set of molecular markers more easily. This will assist your breeding program. Frex: we are have species X and are crossing a oil-seed variety with a fiber variety and then tracking the offspring of the population to see which gives us the best fiber traits... those are then bred to each other for a few generations before we cross it either back to one of the parent varieties or into a new variety. 4) There have been a lot of advances in high though-put molecular marker identification, and our ability to look at populations. Typically in the 1990's there were ways to used marker assisted selection but they were low throughput as opposed to what we do now (a mere ~15 years later) - and dealt with simple traits (colour and such). The problem is that you are trying to look at pretty complex traits (IQ etc etc) which are not inherited as directly. So placed in the 1990's your ability to use any molecular markers would be significantly decreased. Write off your first generation as not worth pursuing (have the government abandon the project). Have them pick it back up about 10 to 20 years later.... now we can look at 10 thousand genes at a time and find trends among them that possibly correlate to complex traits.... give your government 10 to 20 years to get a good set of markers for the traits, simply by sampling their population.... Then do what we do in the plant world (very morally wrong if you are talking humans, but I assume we are talking about a government program that doesn't consider morals and only outcomes).... Take genetics from those you want to add to your program... take your 100,000 eggs and as much sperm as you need, once they start growing - screen them, get your trait pattern grow the offspring to a point were their testing becomes obvious (which could be really early depending on what trait you are looking for). Collect eggs and sperm (it's already there, you are using a test tube... why wait for them to mature physically) re-breed based on the altered profile (a profile can take as little as a couple weeks to work up if you know what you generally are looking for and everything goes right). Do this over 10 generations... which assuming your trait pattern is correct could take between 20 and 150 years. In the mean time your government scientists are likely researching the genes that showed up in your screens for your traits... and may actually out pace your breeding program, and produce a genetic clone imbued with the trait before your breeding program gets there. Typically breeding programs are used when a transgenic creature can't be produced, or the trait you are going after is a complex multi-gene mess. They are also used as a backup plan because the current public sensibility is that transgenic crops/animals are bad and breeding them is more 'natural'. -Sue --- [EMAIL PROTECTED] wrote: > > I just picked up G: Bio-Tech 4e, with a particular > interest in eugenics for an idea that had struck me. > Anyway, I wanted to field opinions on a > hypothetical situation, particularly those of you > with stronger genetics backgrounds than me. > > Scenario: > > c. 1990, a world power starts a eugenics program > with the idea of cultivating intelligence operatives > and other military specialists via genetic > selection, as well as training from infancy. The > initial primary goal is to develop counter-terrorist > field agents, but those parameters expand as the > program expands over time. > > G: Bio-Tech gives a range for game purposes of > improvements that might be had over each generation > that this eugenics program takes place. By the time > c. 2170 arrives, an ongoing eugenics program would > be in its 7-10th generation in vitro (the overlap > because there would be pairings across a range of > ages based on phenotype attributes-- that is, when > one generation hits 18y/o, pairings are made, but > the same generation might be used 10 years later if > a promising pairing could be made with another > genotype-- this, the second generation even would be > conceived across an extended timeframe-- this would, > of course, multiply over time-- anyway, I digress). > Bio-Tech states that (on the stricter end), 1-2CPs > in improvement might be seen per generation. More > generously, it suggests possibly ONE ADVANTAGE might > be gained per generation. > > I'd like to base this model somewhat realistically, > so am fielding opinions-- what is realistic? > According to my limited sources, such a program > would likely recruit genotypes with high > intelligence and a talent for languages, good 3-D > spatial/hand-eye coordination, toughness/durability, > strong will, and athleticism. In a eugenics > situation, wouldn't you start with the best you > could find? > > For example, I'd get the genotype of Steve Young or > someone like him (former 49ers QB). He's smart-- > got his law degree, and probably has a reasonable > talent for language. Obviously has good hand-eye > coordination, good at 3-d spatial tasks with his > hand-eye coordination (throwing the football), > durable, and athletic (ran a 4.5 40 in his prime). > I might pair him with a similar female genotype. > > Clearly, someone like Steve Young is representative > of that combo of attributes somewhat towards the > pinnacle of what humans have to offer. So my > question is-- even if paired with a similar female > genotype, what would the normal distribution of > genetic pairings (i.e. offspring) look like? Assume > that a eugenics program could artificially fertilize > 100-ish zygotes and grow them in test tubes. Given > that Steve Young and his pair are both at the > rarified high end of the distribution, how much do > their offspring shift the distribution beyond > normal? FrEx, if one pairs > a ST12 DX15 IQ13 HT14 Steve Young with a similar > female genotype, what is the distribution of their > offspring? Yes, some will be Attribute 10 across > the board, but that certainly won't be the average. > What is more "realistic"? Especially as one > sustains a eugenics program over the generations > (and certainly phenotype stability will be one of > the attributes such a program would look for). For > example, I'd be a decent candidate (except that I'm > a wimp) because across the generations of my family, > we've been blessed with above-average smarts and > athleticism--> whereas someone smarter and more > athletic that's an aberration in his genetic lineage > might not be as desirable... > > My inclination is to take the Bio-Tech > recommendations, but use the attributes of my base > genetic stock (that is, the attributes of Steve > Young) as the starting point. One would sift the > best offspring off the top for pairing with other > candidates from the larger gene pool (to avoid > inbreeding) and go again. After 10 generations, I > get Steve Young plus 10cps for attributes/advantages > at a minumum, or Steve Young plus 10 advantages at > best... > > One question that might help-- does anyone know > approximately how many generations it takes for a > beetle species to go from flight to flightless when > placed in an environment that is highly selective > against flight, for example? Does that number even > come close to modeling a similar circumstances in > humans? How quickly would I have almost all 7-foot > humans if we only allowed seven-footers to reproduce > with each other, and nobody else? Is it 1-2 > generations, or 10, or more? I think probably only > 1-2, to be honest. But, that's why I'm asking-- and > a collection of attributes is a lot more > complicated... > > Anyway, thanks for the insights... > > -vk > > > _______________________________________________ > GurpsNet-L mailing list <[email protected]> > http://mail.sjgames.com/mailman/listinfo/gurpsnet-l > _______________________________________________ GurpsNet-L mailing list <[email protected]> http://mail.sjgames.com/mailman/listinfo/gurpsnet-l
