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

that supposedly remained in Africa and the group that supposedly left. That is unexpected under the Out of Africa Model. 4. Phylogenetic relationships between haplogroup founders: Both phylogenies have an asymmetrical, star-like topology. Most of the major haplogroup founders are tightly clustered, suggesting a small initial population that underwent explosive population growth. This is not surprising, as genomic data support this idea in general, but from the chrYdata it is clear that most men in the world are descended from a small number of closely-related individuals. A similar thing can be said of the mitochondrial lineages. Since mutation occurs more or less at random, and since the mutation rate in the Y and mitochondrial chromosomes can be less than one per chromosome per generation, were we to run the clock backward and start the post-Flood dispersion again, we would not necessarily get the exact same tree. However, one would get a similar pattern . The asymmetry in both trees is interesting in that is it so similar. A long branch separates the Eurasian groups from groups more closely-associated with Africa, and then rare African groups form long, spidery branches from that point. Clearly, ancient demographic processes are shaping the genetic landscape, but how much of this is demography and how much of this reflects the pre- and immediately post-Babel population genetics is unknown. 5. The effects of low coverage: The main difficulty with the divergence data comes from the fact that the 1000 Genomes data are low coverage. Low sequence coverage makes it difficult to detect short indels and copy number variations, but any effect is expected to be small since, with only a few exceptions, the types of variation known to exist cover only a limited number of SNVs. And even though a certain number of low frequency variants were expected to be missed, when we compared divergence to the sequence coverage, no trend towards higher divergence with lower coverage was revealed (data not shown). 6. Validation using high-coverage sequencing data: We repeated our methods on the Y chromosomes of the 25 individuals included in the Complete Genomics panel of 69 high-coverage genomes. We built a phylogenetic tree and then did extensive comparisons between the Complete Genomics, 1000 Genomes, and Poznik et al. (2016) data. Among these 25 Y chromosomes, 1000 Genomes detected 4,689 SNVs. Complete Genomics added another 377 in those same regions and an additional 5,010 outside the areas covered by 1000 Genomes (Fig. 4). If we only consider those variable positions detected by both 1000 Genomes and Complete Genomics, there are only 39 points (out of 25 x 4,689 readings, or 0.034%), distributed among 23 genomic locations, where the two data sets contradicted one another. If we assume that Complete Genomics always corrects 1000 Genomes, in about half the cases (12/23) Complete Genomics calls a private allele that 1000 Genomes missed. This is an average of 0.44 missed private alleles per person. But this rate would be much lower in the larger data set of 1,233 Y chromosomes. Visually examining these 23 locations in the larger data set revealed that 15 of these locations are variable within one of the major haplogroups and another two are fixed within a haplogroup (and thus none of these are private alleles in 1000 Genomes). The remaining six locations are complex, with much homoplasy but always with a clear majority allele. Obviously, we have reached the limits of current sequencing technology, but the expected number of false positives in the 1000 Genomes Y chromosome SNV dataset is less than 1 per person, on average. Of the 377 places where Complete Genomics called a variable allele that was missed by 1000 Genomes, 75.9% identified a private allele. This is not surprising when you consider that 1000 Genomes is expected to miss a greater percentage of rare variation due to low average sequence coverage. However, false negatives are still much less than one per person. 7. Divergence of individuals from their haplogroup founder: The time elapsed between the ancestor of any haplogroup and the modern members of that group, by definition, is always the same for all individuals in the group. Fig. 5 shows the average distance from each Y chromosome haplogroup founder to the members of that specific haplogroup. We were initially surprised to observe that Carter et al. ◀ Y Chromosome Noah and mitochondrial Eve ▶ 2018 ICC 138 Figure 4. SNV locations in the 1000 Genomes (blue) and Complete Genomics (red) datasets. The Y chromosome centromere starts at about nucleotide position 10.3 million, which accounts for the large gap there. The long heterochromatic area (beyond nucleotide position 30 million) was not sequenced in either study. L1 L5 L2 L4/6 L3 M I/S N R L1  5 14 18 20 20 22 23 25 L5 5  9 13 15 15 17 18 20 L2 14 9  4 6 6 10 11 13 L4/6 18 13 4  2 2 6 7 9 L3 20 15 6 2  0 4 5 7 M 20 15 6 2 0  4 5 7 I/S 22 17 10 6 4 4  1 3 N 23 18 11 7 5 5 1  2 R 25 20 13 9 7 7 3 2  Table 2. Distances between all mitochondrial chromosome haplogroup ancestors. There are fewer groups here than in Fig. 2 because we combined the members of macrohaplogroup R (they were found to have identical mitochondrial ancestors). Only a small number of mutations accumulated in the human population prior to our spreading out across the world.