From: me22over7@gmail.com   
      
   On 29/08/2025 1:48 am, RonO wrote:   
      
   ...   
      
   >> My question then, is it reasonable, accurate and in good faith to   
   >> characterise this as "science doesn't know X therefore God"?   
   >>   
   > Yes it is.  God-of-the-gaps is just that no matter if the gaps are never   
   > going to be filled.  No one expects to figure out how life actually   
   > arose on this planet.  The best that we can do without a time machine is   
   > figure out the most likely way that life arose, and it could have   
   > actually arisen in a much more unlikely way.  The first organic   
   > conglomerate that produced the first self replicating molecules might   
   > have been created by a random asteroid impact instead of condensing in   
   > some crevice somewhere.   
   >   
   > Gaps like what caused the Big Bang may never be filled due to the   
   > singularity that we can't get past to see what existed before.  Denton   
   > is happy with using this gap because he can never be demonstrated to be   
   > wrong.   
   >   
   > I was a genetics major as an undergraduate at Berkeley, and all the   
   > geneticists told me that genetics was over and that the future was   
   > molecular biology, so I did my PhD in molecular genetics.  During my   
   > graduate studies tools for genomic analysis like PCR and microsatellite   
   > genetic markers were created, and they had the promise of allowing the   
   > unsolved issues in genetics to be answered.  I went back into genetics   
   > because the new technology was opening up a new frontier in genetics.   
   > There was finally the possibility that things like why the infinite   
   > allele model worked so well in quantitative genetics.  Quantitative   
   > genetic analysis was well worked out, but we didn't understand why it   
   > worked.  I started my first post doc on looking for quantitative trait   
   > loci (QTL) in 1992.  I retired in 2024 and we still do not know why the   
   > infinite allele model works for quantitative genetic analysis.  We can   
   > sequence whole genomes now, but somethings remain a mystery.  You do not   
   > see the ID perps claiming that their designer is responsible for the   
   > success of the infinite allele model in quantitative genetics.   
   >   
   > My guess for why it works is that due to the fall off in increased   
   > accuracy with the inclusion of more past generations of data (about 3   
   > past generations are enough to use for things like BLUP (best linear   
   > unbiased prediction) is that linkage is the reason for the infinite   
   > allele model working so well.  Most of it may be apparent linkage and   
   > not actual linkage on the same chromosome.  We have found that an LD   
   > (linkage disequalibrium) of only 0.3 within a population is sufficient   
   > to use markers spaced along the genome to improve the accuracy of   
   > predicting the best breeding values for individuals.  All the alleles   
   > segregating in the genome have a minimum linkage of 0.5 (you inherit   
   > half of one parent's genome) in the first cross.  Closely linked markers   
   > have a linkage close to 1.0.  This means that the whole genome can be   
   > considered to have a quantitative genetic value due to hundred or   
   > thousands of genetic variants in that genome, and these variants can be   
   > inherited together just by chance if they occur in the same individual   
   > even if they are not linked.  This would create pseudo haplotypes based   
   > on the whole genome, and not just haplotypes of the same chromosomes. In   
   > the first cross half the QTL segregate together in any individual. My   
   > guess is that this would create close enough to an infinite number of   
   > pseudo haplotypes of unlinked QTL.  Half the sires genome is inherited   
   > in the first cross, 25% in the next cross, but if we are talking genome   
   > equivalents (and things that sire is homozygous for) then all the   
   > homozygous alleles (1 genome equivalent) is inherited among the progeny   
   > of the first cross, half a genome equivalent in the second cross and 25%   
   > in the third cross.  Statistically these pseudo haplotypes could persist   
   > for multiple generations.  Each individual starts with a new set of   
   > pseudo haplotypes that add to the haplotypes due to actual linkage, so   
   > you never see all the "alleles" possible in a single generation.  There   
   > would be pretty much an infinite number of them.   
   >   
   > Ron Okimoto   
   >   
      
   How is retirement treating you?   
      
   Based on your experience as described above (credit where due btw) I'd   
   be interested in your comments on mathematical vs simulation vs observe   
   and infer approaches the study of population genetics.   
      
   I've mentioned here previously a partially completed computer program to   
   simulate a population of ~10,000 "genomes", subject to sexual   
   reproduction, with chromosomes, recombination and crossover, controlled   
   randomisation, and mutations using various selection coefficient   
   profiles, etc. I've been curious to attempt to explore fixation, viable   
   selection coefficient distributions, genetic load and so on.   
      
   Inconclusive so far, but the exercise has been an impetus to try and   
   understand some of the principles involved. I plan to get back to it,   
   but further study of pop gen first would help verify my assumptions and   
   modelling.   
      
   My initial approach was to start with a supposed selection coefficient   
      
   [continued in next message]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)