supraHex (version 1.10.0)

visTreeBootstrap: Function to build and visualise the bootstrapped tree

Description

visTreeBootstrap is supposed to build the tree, perform bootstrap analysis and visualise the bootstrapped tree. It returns an object of class "phylo". For easy downstream analysis, the bootstrapped tree is rerooted either at the internal node with the miminum bootstrap/confidence value or at any customised internal node.

Usage

visTreeBootstrap(data, algorithm = c("nj", "fastme.ols", "fastme.bal"),
metric = c("euclidean", "pearson", "spearman", "cos", "manhattan",
"kendall", "mi"), num.bootstrap = 100, consensus = FALSE,
consensus.majority = 0.5, reroot = "min.bootstrap",
plot.phylo.arg = NULL, nodelabels.arg = NULL, visTree = T,
verbose = T, ...)

Arguments

data

an input data matrix used to build the tree. The built tree describes the relationships between rows of input matrix

algorithm

the tree-building algorithm. It can be one of "nj" for the neighbor-joining tree estimation, "fastme.ols" for the minimum evolution algorithm with ordinary least-squares (OLS) fitting of a metric to a tree structure, and "fastme.bal" for the minimum evolution algorithm under a balanced (BAL) weighting scheme

metric

distance metric used to calculate a distance matrix between rows of input matrix. It can be: "pearson" for pearson correlation, "spearman" for spearman rank correlation, "kendall" for kendall tau rank correlation, "euclidean" for euclidean distance, "manhattan" for cityblock distance, "cos" for cosine similarity, "mi" for mutual information

num.bootstrap

an integer specifying the number of bootstrap replicates

consensus

logical to indicate whether to return the consensus tree. By default, it sets to false for not doing so. Note: if true, there will be no visualisation of the bootstrapped tree

consensus.majority

a numeric value between 0.5 and 1 (or between 50 and 100) giving the proportion for a clade to be represented in the consensus tree

reroot

determines if and how the bootstrapped tree should be rerooted. By default, it is "min.bootstrap", which implies that the bootstrapped tree will be rerooted at the internal node with the miminum bootstrap/confidence value. If it is an integer between 1 and the number of internal nodes, the tree will be rerooted at the internal node with this index value

plot.phylo.arg

a list of main parameters used in the function "ape::plot.phylo" http://www.inside-r.org/packages/cran/ape/docs/plot.phylo. See 'Note' below for details on the parameters

nodelabels.arg

a list of main parameters used in the function "ape::nodelabels" http://www.inside-r.org/packages/cran/ape/docs/nodelabels. See 'Note' below for details on the parameters

visTree

logical to indicate whether the bootstrap tree will be visualised. By default, it sets to true for display. Note, the consensus tree can not be enabled for visualisation

verbose

logical to indicate whether the messages will be displayed in the screen. By default, it sets to true for display

...

additional "ape::plot.phylo" parameters

Value

an object of class "phylo". It can return a bootstrapped tree or a consensus tree (if enabled): When a bootstrapped tree is returned (also visualised by default), the "phylo" object has a list with following components:

• Nnode: the number of internal nodes
• node.label: the labels for internal nodes. Here, each internal node is associated with the bootstrap value
• tip.label: the lables for tip nodes. Tip labels come from the row names of the input matrix, but are not necessarily the same order as they appear in the input matrix
• edge: a two-column matrix describing the links between tree nodes (including internal and tip nodes)
• edge.length: a vector indicating the edge length in the 'edge'
• Note: the tree structure is indexed with 1:Ntip for tip nodes, and ($Ntip$+1):($Ntip$+$Nnode$) for internal nodes, where $Ntip$ is the number of tip nodes and $Nnode$ for the number of internal nodes. Moreover, $nrow(data)=Ntip=Nnode-2$.

When a consensus tree is returned (no visualisation), the "phylo" object has a list with following components:

• Nnode: the number of internal nodes
• tip.label: the lables for tip nodes. Tip labels come from the row names of the input matrix, but are not necessarily the same order as they appear in the input matrix
• edge: a two-column matrix describing the links between tree nodes (including internal and tip nodes)

See Also

visTreeBootstrap

Examples

# 1) generate an iid normal random matrix of 100x10
data <- matrix( rnorm(100*10,mean=0,sd=1), nrow=100, ncol=10)
colnames(data) <- paste(rep('S',10), seq(1:10), sep="")
data <- t(data)

# 2) build neighbor-joining tree with bootstrap values and visualise it by default
visTreeBootstrap(data)

# 3) only display those internal nodes with bootstrap values > 30
# 3a) generate the bootstrapped tree (without visualisation)
tree_bs <- visTreeBootstrap(data, visTree=FALSE)
# 3b) look at the bootstrap values and ordered row names of input matrix
# the bootstrap values
tree_bs$node.label
# ordered row names of input matrix
tree_bs$tip.label
# 3c) determine internal nodes that should be displayed
Ntip <- length(tree_bs$tip.label) # number of tip nodes
Nnode <- length(tree_bs$node.label) # number of internal nodes
flag <- as.numeric(tree_bs$node.label) > 30
text <- tree_bs$node.label[flag]
node <- Ntip + (1:Nnode)[flag]
visTreeBootstrap(data, nodelabels.arg=list(text=text,node=node))

# 4) obtain the consensus tree
tree_cons <- visTreeBootstrap(data, consensus=TRUE, num.bootstrap=10)