Data from: Exploiting gene families for phylogenomic analysis of myzostomid transcriptome data

BACKGROUND: In trying to understand the evolutionary relationships of organisms, the current flood of sequence data offers great opportunities, but also reveals new challenges with regard to data quality, the selection of data for subsequent analysis, and the automation of steps that were once done manually for single-gene analyses. Even though genome or transcriptome data is available for representatives of most bilaterian phyla, some enigmatic taxa still have an uncertain position in the animal tree of life. This is especially true for myzostomids, a group of symbiotic (or parasitic) protostomes that are either placed with annelids or flatworms. METHODOLOGY: Based on similarity criteria, Illumina-based transcriptome sequences of one myzostomid were compared to protein sequences of one additional myzostomid and 29 reference Metazoa and clustered into gene families. These families were then used to investigate the phylogenetic position of Myzostomida using different approaches: Alignments of 989 sequence families were concatenated, and the resulting superalignment was analyzed under a Maximum Likelihood criterion. We also used all 1,878 gene trees with at least one myzostomid sequence for a supertree approach: the individual gene trees were computed and then reconciled into a species tree using gene tree parsimony. CONCLUSIONS: Superalignments require strictly orthologous genes, and both the gene selection and the widely varying amount of data available for different taxa in our dataset may cause anomalous placements and low bootstrap support. In contrast, gene tree parsimony is designed to accommodate multilocus gene families and therefore allows a much more comprehensive data set to be analyzed. Results of this supertree approach showed a well-resolved phylogeny, in which myzostomids were part of the annelid radiation, and major bilaterian taxa were found to be monophyletic.

背景：在探究生物演化关系的过程中，当前海量的序列数据既带来了难得的发展机遇，也在数据质量、后续分析的数据筛选，以及曾需手动完成的单基因分析流程自动化等方面提出了全新挑战。尽管多数两侧对称动物门（bilaterian phyla）的代表类群已具备基因组或转录组数据，但部分神秘分类群（enigmatic taxa）在动物生命之树上的分类位置仍未明确。其中尤以吸口虫类（Myzostomida）为甚：该类群为一类共生（或寄生）的原口动物（protostomes），其分类地位曾被归入环节动物（annelids）或扁形动物（flatworms）。
研究方法：基于相似性标准，将1个吸口虫类的基于Illumina测序的转录组序列，与另外1个吸口虫类以及29个参照后生动物（Metazoa）的蛋白质序列进行比对，并聚类为基因家族（gene families）。随后利用这些基因家族开展吸口虫类系统发育位置的探究，采用两种不同分析策略：其一为拼接策略，将989个序列家族的比对结果拼接为超级比对矩阵（superalignment），并基于最大似然法（Maximum Likelihood）标准进行系统发育分析；其二为超树（supertree）策略，纳入所有包含至少1条吸口虫类序列的1878个基因树（gene trees）：首先计算各独立基因树，再通过基因树简约法（gene tree parsimony）将其整合为物种树（species tree）。
结论：超级比对矩阵分析要求严格的直系同源基因（orthologous genes），但本数据集内的基因筛选环节，以及不同分类群所获数据量的显著差异，均可能导致异常分类位置推断与较低的自展支持率（bootstrap support）。与之相对，基因树简约法专为适配多座位基因家族而设计，因此可支持更为全面的数据集分析。本研究采用超树策略得到的结果，构建了分辨率良好的系统发育树，其中吸口虫类隶属于环节动物辐射演化支，且主要两侧对称动物类群均呈现单系性（monophyletic）。