**Revisiting the Coffea–Psilanthus relationship using nuclear SNP phylogenomics**  

The fusion of the Psilanthus genus into Coffea proposed by Davis et al. (2011) was primarily based on phylogenetic analyses using nuclear ribosomal internal transcribed spacer (ITS) and a limited number of chloroplast markers, combined with overlapping morphological characters.  
However, the arguments used are because thebecause of the problematic: plesiomorphies are suggested to be apomorphies, low phylogenetic resolution of these markers used,and the potential influence of incomplete lineage sorting (ILS) and introgression remained unresolved. 
Here, we reassess the evolutionary relationships between Coffea and Psilanthus using a genome-wide nuclear dataset comprising both 28,800 SNPs primarily established by Hamon et al. 2017 and high throughput resequencing across more than xx species representing the major diversity of both groups. Nuclear phylogenomic analyses reveal a strong and consistent separation between Coffea and Psilanthus, with an additional strong divergence between African and Asian Psilanthus lineages. In contrast, complete chloroplast phylogenies show extensive incongruence with the nuclear topology and provide limited resolution, supporting previous evidence of chloroplast capture and/or incomplete lineage sorting. Coffea rhamnifolia, previously identified as morphologically and phylogenetically ambiguous, occupies a basal position relative to the Psilanthus clade in nuclear analyses, suggesting a key role in the early diversification of these lineages.
Our results highlight the importance of genome-scale nuclear data for resolving complex tropical plant radiations.  
It also provides a new framework to reconsider the taxonomy and the evolutionary history of coffee relatives.
This Git offers the script used for this reasearch.