Supplementary materials for: Improving detection in large-scale gene family expansions and associated phylogenomic inferences in Ericales and Caryophyllales
收藏资源简介:
Ancient whole-genome duplications (WGDs) have shaped flowering plant evolution, but their placement is sometimes uncertain, especially in clades primarily known from transcriptomes. In such datasets, peaks in synonymous substitutions among paralogs (Ks) and phylogenomic gene-duplication mapping can identify large-scale gene family expansions (GFEs), but these approaches may disagree and, by themselves, cannot distinguish WGD from segmental duplication or other mechanisms. These discrepancies may stem from biological processes such as diploidization, hybridization, and incomplete lineage sorting, or from systematic biases introduced by hidden paralogy that current pipelines fail to address. Here, we present H2O (homolog-to-ortholog), a workflow that integrates Ks distributions with tree-based gene duplication and orthology inference to improve the detection and placement of transcriptome-inferred GFEs while reducing errors due to hidden paralogy. We evaluated this approach using datasets from Ericales and five Caryophyllales subgroups, refining large-scale GFE placements and comparing phylogenomic performance with ASTRAL-Pro and Yang and Smith (2014). Across six clades, this integrated approach improved confidence in GFE placement and, under conservative filtering, reduced associated gene-tree discordance. Reduced discordance comes at the cost of substantial data loss. Our results demonstrate that Ks values are strongly influenced by life history, with herbaceous lineages exhibiting shifted distributions relative to woody relatives due to faster molecular rates. Furthermore, we found that inferred gene-family retractions, measured as duplicate-copy absences, were often concentrated on branches preceding major named clades. These results highlight the need to treat transcriptome-inferred GFEs cautiously, integrate multiple lines of evidence, and explicitly account for hidden paralogy in plant phylogenomics.



