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

phylogenetic methods (including parsimony, maximum likelihood, and Bayesian approaches) yield slightly different results (Hanson- Smith et al. 2010; Groussin et al. 2015). Thus, it is expected that our ancestral reconstruction methods will produce results slightly different from other methods. Yet, as long as all sequences are being compared to the correct ancestor, as long as that ancestor is not biased in the direction of some descendant sequences over others (as would occur in a strict consensus model, for example), and as long as enough mutations have occurred in each lineage to produce a statistically-robust average, ancestral reconstruction should produce accurate results. The distances of each sequence to its group ancestor was calculated and group averages and standard deviations were tabulated. Sequence manipulation and most calculations were performed in Perl or MEGA. We used Tukey’s Multiple Comparisons of Means to calculate family-wise 95% confidence intervals for all pair-wise divergence differences among all sequences compared to each ancestral sequence. Once we saw that the rates of mutation accumulation were not identical in all lineages, we decided to explore why this might be true. We obtained the sequence data from the 50,000th-generation of E. coli (Tenaillon et al. 2016) grown in the Long-Term Evolution Experiment (LTEE) pioneered by Richard Lenski. We compared the relative proportion of each of the 12 SNV types in the bacterial chromosomes and within human chromosome 22. RESULTS 1. A comprehensive phylogeny for the Y and mitochondrial chromosomes: The unrooted neighbor-joining phylogenic trees for the Y and mitochondrial chromosomes are show in (Figs. 1–3). There are several interesting things that can be seen in these images. First, there is always a clear, central starburst pattern. Since most new mutations are lost to drift with time (Rupe and Sanford 2013), the only way to capture a pattern like this is if the human population expanded extremely rapidly and/or if it had an exceptionally high mutation rate at an earlier period of its history. Comparing the natural groupings revealed in the phylogenetic trees to the nearest- neighbor data allowed us to identity 11 major haplotypes for chrY and 16 major haplogroups for chrM. Some of these were collapsed into larger groups when ancestral reconstruction revealed that they had identical ancestors. 2. Ancestral sequences for each Y chromosome haplogroup: Applying the first-pass tests (cases A, B, and C in the ancestral reconstruction methods described above) led to unambiguous ancestral predictions for 98.3% of all variable positions among all Y chromosome haplogroups. This is similar to the reported ambiguity found in the mitochondrial dataset of Carter et al. (2008). The data were complicated by the presence of multiple apparent homoplasies. These are mutations that occur in parallel in independent lineages, including hundreds of locations in the mitochondrial data and thousands of locations in the Y chromosome data. Nearly all were resolved using the second-pass test (the special cases mentioned in Methods). Especially important was the removal of unique within- group alleles (i.e., the only reason the homoplasy existed was that an allele associated with a major phylogenetic branch point also appeared in a single individual in an unrelated group). This either revealed many sequencing errors, which is unlikely, or thousands of examples of repeating mutations or gene conversion events at the same locations in disconnected lineages, which has significant implications for phylogenetics. Sorting and visual examination of the 1000 Genomes data showed that there were no ancestral allele calls that contradicted the main branches on the standard Y chromosome phylogenetic tree (c.f. Scozzari et al. 2014). However, several ancestral mitochondrial allele calls were different from the most recent phylogenetic work (e.g., Behar et al. 2012). There was so much recurrent mutation at several places that the ancestral allele was uncertain: either there was not a clear consensus, homoplasy existed in the majority of branches, or the pattern made no sense compared to the overall tree. Thus, contrary to Behar et al. (2012), we found no differences in the ancestral sequence of haplogroups L4 and L6, and haplogroup I/S was removed from macrohaplogroup N by a single mutation (at position 10398), whereas all of the other macrohaplogroup R branches split off directly from a common node after that. 3. Differences among haplogroup founders: The distance matrix for the Y chromosome haplogroup founders is given in Table 1. It reveals three large clusters of closely-relatedYchromosome groups. The distance matrix for the mitochondrial haplogroup founders is given in Table 2. It shows that multiple major lineages (e.g., B, F, H/V/R, J/T, and U/K) branch off directly and simultaneously (from a tree-building perspective) from a single ancestral sequence (in this case, the founder of macrohaplogroup R). The presence of multiple early women who were both closely related and who were the founders of large proportions of the current world population is surprising, to say the least, unless one is considering biblical history. If all haplogroups branched off from within a population of ~10,000 individuals, founders should essentially never be closely related. It is also important to note that the ancestor of the “Out of Africa” clade (L3) is identical to the ancestor of macrohaplogroup M. There was no discernable time between the rise of the group Carter et al. ◀ Y Chromosome Noah and mitochondrial Eve ▶ 2018 ICC 137 Table 1. Distances between all Y chromosome haplogroup ancestors. There are several major groupings evident in this table (shaded areas), including a group that includes the closely-related ancestors of L/T, N/O, and Q/R (macrohaplogroup K), as well as the ancestors of G, H and I/J, and the ancestors of D/E and C. These three groups represent most of the Y chromosome lineages in the world. Actually, macrohaplogroup K has that distinction by itself, but the other groups still represent a significant percentage of world ancestry. Thus, the majority of worldwide Y chro- mosome haplogroups immediately descend from one of three macrohap- logroup ancestors. A1 B C D/E G H I/J L/T N/O Q/R A1 422 653 649 816 817 824 840 841 841 B 422 233 229 396 397 404 420 421 421 C 653 233 4 163 164 171 187 188 188 D/E 649 229 4 167 168 175 191 192 192 G 816 396 163 167 1 8 24 25 25 H 817 397 164 168 1 7 23 24 24 I/J 824 404 171 175 8 7 16 17 17 L/T 840 420 187 191 24 23 16 1 1 N/O 841 421 188 192 25 24 17 1 0 Q/R 841 421 188 192 25 24 17 1 0