Wheeler W C, Gatesy J, DeSalle R
Department of Invertebrates, American Museum of Natural History, New York, New York 10024-5192, USA.
Mol Phylogenet Evol. 1995 Mar;4(1):1-9. doi: 10.1006/mpev.1995.1001.
The process of multiple sequence alignment provides homology statements for the phylogenetic analysis of molecular data. Unfortunately, multiple alignments are frequently nonunique. Two sources of these multiple alignments are analysis based on different sets of alignment parameter values (gap:change cost ratios) and nonunique equally costly alignments based on a single set of alignment parameters. By "eliding" these individual alignments into a single grand alignment, phylogeny that is weighted toward those positions that align more consistently can be reconstructed. Positions that show greater variation among alignments will be relatively downweighted. The technique results in a weighting procedure that is a posteriori and based on the evidence established from the original sequence alignments.
多序列比对过程为分子数据的系统发育分析提供了同源性声明。不幸的是,多序列比对结果常常不唯一。这些多序列比对结果不唯一的两个原因是:基于不同比对参数值集(空位:变化成本比率)的分析,以及基于单一比对参数集的同等成本的不唯一比对。通过将这些单独的比对“合并”为一个总的比对,可以重建更倾向于那些比对更一致的位置的系统发育。在各比对之间显示出更大变异的位置将被相对加权。该技术产生了一种基于后验且基于原始序列比对所建立证据的加权程序。