Performs tree inference under "RAxML"
for aligned fasta sequences in
a given folder (default is "2.Alignments"
). Note that you need at
least two gene regions to run a partitioned analysis.
Usage
tree.raxml(
folder = "2.Alignments",
FilePatterns = "Masked_",
raxml_exec = "raxmlHPC",
Bootstrap = 100,
outgroup,
partitioned = FALSE,
...
)
Arguments
- folder
Name of the folder where the sequences to align are stored (character).
- FilePatterns
A string that is common to all the target files in the relevant folder (character). Note that this argument can be set to
"NULL"
if no specific pattern wants to be analized.- raxml_exec
Where to find
"RAxML"
or how to run it from the console? (string).- Bootstrap
Number of bootstrap replicates (numeric).
- outgroup
A single string of comma-separated tip labels to be used as outgroup in
"RAxML"
See"RAxML"
documentation for more details (character).- partitioned
Whether analyses should be partitioned by gene (Logical).
- ...
Arguments passed to
"ips::raxml"
.
Examples
if (FALSE) {
sq.retrieve.direct(
clades = c("Felis", "Vulpes", "Phoca"),
species = "Manis_pentadactyla",
genes = c("ADORA3", "CYTB")
)
sq.curate(
filterTaxonomicCriteria = "Felis|Vulpes|Phoca|Manis",
kingdom = "animals", folder = "0.Sequences"
)
sq.aln(folder = "1.CuratedSequences")
tree.raxml(
folder = "2.Alignments", FilePatterns = "Masked",
raxml_exec = "raxmlHPC", Bootstrap = 100,
outgroup = "Manis_pentadactyla"
)
}