As discussed above, the sites of mixed alleles in
the PCR product direct sequencing are considered true polymorphisms
because errors introduced by Taq polymerase are
randomly distributed in each individual product, and the
proportion of the random mutation to the initial template
population is small. Consequently, the major sequencing signal
at any one nucleotide will be from the original DNA templates
with no mutations. Among the 57 sites detected from chicken
ZZ24 and HC2, 35 (61.4%) had mixed alleles in PCR product
direct sequencing, which was within the range of the estimated
rate of true differences (,80%) using the enzyme error rate. To
answer the second question, we removed the polymorphic sites in
the individual cloned coxIII sequences not matched with the 35
sites in the sequences of PCR product direct sequencing from
ZZ24 and HC2, creating individual sequences with polymorphic
sites verified by PCR product direct sequencing. We then
performed hierarchical clustering and estimated the genetic
distances between the cloned sequences. We obtained a dendrogram
similar to that in Figure 4A