Table 1.
Species considered in this study for chloroplast genomes analyses.
Herbarium codes after Holmgren et al. (1990). Germplasm source BR (Botanic Gardens of Belgium), BRC (Biological Resources Center, Reunion), CBI (Coffee Board of India), K (Royal Botanic Gardens, Kew (UK), KCRS (Kianjavato Coffee Research Station (Madagascar), P (Natural History Museum, Paris (France)).
Fig 1.
Annotation differences from Dogma and GeSeq using C. arabica Cp genome as study case.
The Cp genome, split into two parts, is described as an ordered list of predicted coding sequences, leading to two black dotted lines for each part, surrounded by gene names written in blue. In each part, the upper line entitled C. arabica corresponds to the Dogma annotations, while the lower line is for GeSeq. A red edge indicates paralogous genes within a given genome, while a blue line is drawn when predicted sequences are similar between the two annotations.
Fig 2.
Evolution of gene content based on Dogma annotations.
Each node contains the genome pan genes that are missing in the species under consideration (at the level of leaves, for species currently living, and at the level of internal nodes, concerning their ancestors). Gene insertion or deletion events are indicated at the branch level.
Fig 3.
Maximum likelihood nuclear phylogeny of coffee trees.
The molecular tree was based on 22,800 SNPs as described in [18] and a larger set of species (see S1 Table). Branch lengths are proportional to inferred nucleotide substitutions. The values at the nodes indicate the bootstrap support of each branch in percent. The main clades are shaded, i.e., blue for Psilanthus, pink for Coffea and among this latter group the main geographic areas are shaded pale green for Africa, yellow for Mascarenes and pale grey for Madagascar and Comoros (yellowish branch).
Fig 4.
Nuclear phylogenetic network representation.
The NeighborNet method of the Maximum likelihood nuclear phylogeny of coffee trees presented in Fig 3 was used to obtain the phylogenetic network.
Fig 5.
Maternal phylogenetic relationships.
The ML phylogenetic tree was constructed based on a set of 16 coffee trees representatives of the main geographic clades, and whole chloroplast sequences study using Emmenopterys henryi as outgroup. Psilanthus (in blue) and Coffea (in red) species clustered into two well-supported clades with the exception of C. canephora belonging to the Psilanthus clade.
Fig 6.
Cp genome gene contents for some pairwise comparisons based on GeSeq annotations.
Each of the five images compares the two genomes of sister species in phylogeny. Each genome is represented as a dotted line, each dot being a gene (the dot is labeled by its name). A red curve indicates paralogy between genes inside a given genome, while a blue line is for gene orthology between two distinct genomes.
Fig 7.
Gentianales and its neighboring orders phylogenetic relationships.
The tree is based on the taxonomy and data available on the NCBI website.
Fig 8.
Gene mutation rates per species among each clade.
The gene mutation rate between a species and its ancestor for a given gene is equal to the Levenshtein distance between the translated gene sequences of the specie and the ancestor divided by the length of their alignment. (A) Gene mutation rates between the species in clade 1 (Psilanthus) and the reconstructed ancestor of the clade. (B) Gene mutation rates between the species in clade 2 (Coffea) and the reconstructed ancestor of the clade.
Table 2.
Number of species in which one or several genes gene are different from the ancestral version.
Genes have been ordered according to their variability, from highly to less variable and according to the number of species in which these genes are different from the ancestral version. For instance, all the species of Clade 1 (7) have rpl16 and rps16 different from their ancestor.
Fig 9.
Phylogenetic tree showing the number of mutated genes per species.
Phylogenetic tree with the number of mutated genes per species when compared to the ancestor (numbers next to the species names) of the clade (A)- Clade 1 (Psilanthus); (B)- Clade 2 (Coffea). Branch values are bootstraps according to the maximum likelihood phylogenetic reconstruction. A suffix has been added to provide unique names to ancestral nodes.