The Proceedings of the Eighth International Conference on Creationism (2018)

discover which parameters settings, if any, might allow a highly heterozygous first couple to generate allele distributions similar to the currently observed human allele frequency distributions. Since each one of the 88 original autosomes in Eden could have carried its own unique set of designed variants, normal chromosomal recombination and segregation could have generated a large number of genotypes in the second generation. In each succeeding generation allelic diversity would increase due to newly arising mutations and further recombination. Our simulations required the creation and tracking of two very different types of genetic variation. The first type was the classically understood mutational allele , and the second type was the designed allele . Mutational alleles would arise essentially as word-processing errors in the genome. This type of mutational allele would always arise as a rare variant. Mutations are always occurring, and mutation count per individual consistently increases in number. From its inception, Mendel has always tracked each new mutation and each mutational allele. To simulate newly arising mutational alleles, we only had to specify the population’s mutation rate and the effect of each mutation on fitness. Our default mutation rate was 100 mutations per person per generation. Our default mutational effect was “near-zero” (i.e., there was essentially no selection happening, all mutational alleles would be drifting). In addition to mutational alleles, we simulated initial genetic variants that were created as designed allelic pairs , wherein each allele in a pair had its own designed function. Designed allele pairs would be present at the beginning of a Mendel run. To simulate this model, we had to create within Mendel a new computational function which establishes and tracks designed alleles. This new function allows the specification of: a) the number of designed allele pairs and their locations; b) the ratio of the paired alleles (1:1 or 1:3); and c) the fitness effect per pair (pairs are normally given equal but opposite fitness effects). Under the heterozygous Adam and Eve model, there would be just four copies of each chromosome in Eden, and so every designed allele pair would have a ratio of either 50/50 or 25/75 (so all initial allele frequencies for the designed alleles would be either 0.25 or 0.50 or 0.75). For most experiments, the magnitude of the fitness effects was always “near- neutral” (no effective selection). 4. Examining the Designed Gametes Model We examined the logical outcome that would arise if God individually designed each of the gametes (more accurately the gametogonia ) within Eden, with each gamete (or gametogonium) potentially having its own unique genotype. We tested to see if this could possibly generate the allele frequencies observed today. The logic of this analysis is described in the Results section. We first explain that two designed people could have millions of individually designed gametogonia, and that these diverse gametogonia could represent a gene pool essentially equivalent to the gene pool of a large human population. We then illustrate this using numerical simulations. We initially simulated 50 offspring that carried designed alleles from a first couple, which would have been transmitted through 100 genetically independent gametes (50 sperm and 50 eggs). Mendel then tracked the initial designed alleles, plus accumulating mutational alleles, though a 200-generation biblical framework (including population growth, a 6-person bottleneck in generation 9, and re-growth up to a pre-set maximum population size). 5. Complexities of plotting allele frequencies from simulations that include designed alleles As stated above, all new mutations begin as very rare alleles. However, following the standard convention, we do not normally plot alleles with a frequency less than 1.0%. Although these rare alleles account for most of the allelic diversity, we tally, but do not plot the very rare alleles. Instead, in our allele frequency plots the first (left-most) bin tallies the number of alleles with a frequency of 1–2%, the next bin tallies alleles with a frequency of 2–3%, etc. There are numerous practical reasons for this: a) detection of very rare alleles in this first bin is very sensitive to sampling size and so can fluctuate wildly; b) this first bin incorporates all DNA sequencing errors; c) this first bin is usually so large that it severely distorts the scaling of any allele frequency plot. We will revisit the importance of this “invisible bin” in the Discussion section. Another major data plotting issue involves the question of whether we should plot allele frequencies from 1% to 50%, or allele frequencies from 1% to 99%. It is normally assumed that all alleles arise via random mutations, so it is assumed that there is an “original” (ancestral) allele and a “mutant” (derived) allele. It is usually also assumed that the original allele is the one most frequently observed (the major allele), and that the mutant allele is rare (the minor allele). Thus, allele frequency plots normally only show the minor allele (i.e., only allele frequencies between 1% and 50% are plotted), while the major allele is simply assumed. In our simulations, we actually know which alleles are original and which are derived by mutation (when scientists look at real human allele frequency distribution data, they cannot know which allele was the “original” since allele frequencies change over time). The normal convention is that only the minor alleles are plotted. Since for every minor allele at frequency f there exists a major allele at a frequency 1 – f , if the major allele distribution was also plotted, it would appear as a mirror image of the minor allele distribution, making it redundant. For designed allele experiments, we always tally all alleles, but usually only plot the minor alleles (1–50%). Only in a few special cases do we plot all alleles (1–99%). When discussing designed alleles and their allele frequency distributions we again need to clarify our terminology. When we simulate designed alleles, we cannot realistically adopt the terms “ancestral” or “derived”. Likewise, when we specify designed allele pairs, both of the contrasting alleles will often start with a frequency of 50%, so we cannot initially define the “major” or the “minor” allele. However, genetic drift will quickly “break the tie”, at which point we can empirically classify the less abundant allele as the minor allele (as genetic drift continues, the major and minor alleles can “flip” over time). 6. Details of simulations of evolutionary populations We simulated evolutionary human populations where there were no designed alleles. We generally specify 1000 individuals, with 100 new mutations being added per individual per generation. Except where noted, we have made all mutations effectively neutral, as is commonly assumed. This is essential for longer runs over many generations, or the population will go extinct due to slightly Sanford et al. ◀ Designed genetic diversity in Adam and Eve ▶ 2018 ICC 203