Department of Biology, University of Florida, Gainesville, FL 32611, USA.
Pepperwood Way, San Jose, CA 95124, USA.
Biology (Basel). 2013 Mar 13;2(1):419-44. doi: 10.3390/biology2010419.
Insertion/deletion (indel) mutations, which are represented by gaps in multiple sequence alignments, have been used to examine phylogenetic hypotheses for some time. However, most analyses combine gap data with the nucleotide sequences in which they are embedded, probably because most phylogenetic datasets include few gap characters. Here, we report analyses of 12,030 gap characters from an alignment of avian nuclear genes using maximum parsimony (MP) and a simple maximum likelihood (ML) framework. Both trees were similar, and they exhibited almost all of the strongly supported relationships in the nucleotide tree, although neither gap tree supported many relationships that have proven difficult to recover in previous studies. Moreover, independent lines of evidence typically corroborated the nucleotide topology instead of the gap topology when they disagreed, although the number of conflicting nodes with high bootstrap support was limited. Filtering to remove short indels did not substantially reduce homoplasy or reduce conflict. Combined analyses of nucleotides and gaps resulted in the nucleotide topology, but with increased support, suggesting that gap data may prove most useful when analyzed in combination with nucleotide substitutions.
插入/缺失(indel)突变由多序列比对中的间隙表示,已经用于检验一段时间以来的系统发育假说。然而,大多数分析将缺口数据与它们嵌入的核苷酸序列相结合,可能是因为大多数系统发育数据集包含的缺口特征很少。在这里,我们报告了使用最大简约法(MP)和简单最大似然法(ML)框架对鸟类核基因排列中的 12030 个缺口特征进行的分析。两棵树非常相似,它们表现出核苷酸树上几乎所有强烈支持的关系,尽管两棵缺口树都没有支持许多在以前的研究中难以恢复的关系。此外,当不一致时,独立的证据通常支持核苷酸拓扑而不是缺口拓扑,尽管具有高自举支持的冲突节点数量有限。过滤掉短的插入缺失并没有显著减少同形性或减少冲突。核苷酸和缺口的联合分析产生了核苷酸拓扑,但支持度增加,表明当与核苷酸替换一起分析时,缺口数据可能最有用。