This preview shows page 1. Sign up to view the full content.
Unformatted text preview: e sequence data to estimate the age of the common ancestral mtDNA in humans. Their estimate of the 95% confidence interval of the age was 180,000-380,000 years. However, this estimate seems to be less reliable than ours, for three reasons. First, the two-class model that they used is less realistic than our model, as mentioned above. Second, they estimated the proportion fof variable sites from sequence data, but the reliability of their estimate of f is unclear. Actually, the estimate of the number of nucleotide substitutions obtained by using this model is known to be very sensitive to the value of f: Third, the number of human sequences used was much smaller than that used in the present study. In recent years the age of the common ancestral mtDNA in humans has become a controversial issue in relation to the origin of Homo sapiens. Some authors have attempted to estimate the time of the origin of H. sapiens from the age of the common ancestral mtDNA. Theoretically, this attempt is doomed to failure because the age (or the coalescence time) is nothing but a function of long-term effective size of the population (Kingman 1982; Tajima 1983). If the effective size is large, the age of the common ancestor is expected to be large. However, the age has one important evolutionary implication. That is, the age of the common ancestor cannot be smaller than the time at which different populations diverged (Nei 1985 ) . It is now generally agreed that the first major division of human populations occurred between Africans and non-Africans (Nei and Roychoudhury 1982; Cavalli-Sforza et al. 1988; Nei and Ota 199 1). Therefore, our lower-bound estimate of the age of the common ancestral mtDNA suggests that the divergence between Africans and non-Africans occurred at least - 80,000 years ago [or 110,000 years ago, according to Nei’ ( 1992 ) computation]. s This estimate is close to the divergence time ( 115,000 years ago) estimated from data on nuclear genes (Nei and Roychoudhury 1982)) though the latter estimate is very crude. As mentioned earlier, our estimate of the s/v ratio ( 15.7) obtained from analysis of human sequences is very close to that obtained by Vigilant et al. ( 199 1). It is also close to the estimate obtained by our model from the comparison of human and chimpanzee sequences. However, it is considerably smaller than Horai and Hayasaka’ s ( 1990) estimate (37.2) for the hypervariable segment on the 5’ side of the control region. Ward et al. ( 199 1) also examined the s/v ratio for the 5 ‘ -side hypervariable segment, using an independent but small sample, and found no transversions. These 524 Tamura and Nei Table 5 Numbers of Transitional s^ and Transversional 8 Substitutions in the Hypervariable Segments of the mtDNA Control Region
Hypervariable Segment 5’ Side .......... 3’ Side .......... Overall ........ 207 122 8 11 10 Total 218 132 z/‘ /s 18.8 12.2 15.7 329 u 350 NOTE.-The number of nucleotide substitutions was estimated by parsimony -side hypervariable segments are analysis of 95 human sequences. The S- and 3 ’ defined as positions 16028-16362 and 7 l-379, respectively, of Anderson et al.‘ s (1981) sequence. The x2 value for the 2 X 2 table is 0.95 (0.3 < P < 0.5). observations suggest that the 5’ -side hypervariable segment has a higher s/v ratio than does the 3’ side. Because we have more data for both 5’ and 3’ -side segments, we examined this problem. The results obtained (table 5 ) show that the 2/ 6 ratio certainly tends to be higher in the 5 ‘ -side segment than in the 3’ side, but the difference is not statistically different. Therefore, there is no need to treat the two segments separately. However, the a value for the 5’ and 3’ hypervariable segment is not the same...
View Full Document
This note was uploaded on 01/06/2010 for the course NS 2750 taught by Professor Haas&gu during the Spring '08 term at Cornell University (Engineering School).
- Spring '08