Phylotranscriptomic Analysis Based on Coalescence was Less Influenced by the Evolving Rates and the Number of Genes: A Case Study in Ericales

Abstract

Advances in high-throughput sequencing have generated a vast amount of transcriptomic data that are being increasingly used in phylogenetic reconstruction. However, processing the vast datasets for a huge number of genes and even identifying optimal analytical methodology are challenging. Through de novo sequenced and retrieved data from public databases, we identified 221 orthologous protein-coding genes to reconstruct the phylogeny of Ericales, an order characterized by rapid ancient radiation. Seven species representing different families in Ericales were used as in-groups. Both concatenation and coalescence methods yielded the same well-supported topology as previous studies, with only two nodes conflicting with previously reported relationships. The results revealed that a partitioning strategy could improve the traditional concatenation methodology. Rapidly evolving genes negatively affected the concatenation analysis, while slowly evolving genes slightly affected the coalescence analysis. The coalescence methods usually accommodated rate heterogeneity better and required fewer genes to yield well-supported topologies than the concatenation methods with both real and simulated data.

Complete plastid genome sequence ofPrimula sinensis(Primulaceae): structure comparison, sequence variation and evidence foraccDtransfer to nucleus

“…Primulaceae s. l. was placed as sister to the clade that comprising Theaceae, Actinidiaceae, and Ericaceae. The phylogenetic positions of these groups are in agreement with recent studies ( Stevens, 2012 ; Zhang et al, 2015 ).…”

Comparative analyses of three complete Primula mitogenomes with insights into mitogenome size variation in Ericales

“…5 ). Notably, these relationships are broadly consistent with previous studies with multiple fragment combinations (containing plastid, mitogenome, and nuclear genes [ 50 ];), plastomes [ 68 ], and transcriptomes and genomes [ 69 , 70 ].…”

Topological support and data quality can only be assessed through multiple tests in reviewing Blattodea phylogeny

“…AU test) can test alternatives not recovered in the most likely tree (Shimodaira, 2002) but their power is limited by data homogeneity. These multifaceted approaches improve understanding of the dependence of the results on sampling (Narechania et al, 2012;Ware et al, 2008), modelling (Djernaes et al, 2015), alignment bias (Dell'Ampio et al, 2014), or inference method (Djernaes et al, 2015;Zhang et al, 2015).…”

“…The low scores overall and the heterogeneity among pseudoreplicate trees indicates that: more markers are needed to stabilize the tree topology (Narechania et al, 2012) and dilute potential biases; better modelling strategies are needed (e.g. Letsch and Kjer, 2011;Zhang et al, 2015); or missing data patterns should be filled in through new sequencing. In parallel, signal analyses must be conducted to differentiate the markers supporting the relationships from those with highly conflicting signal.…”