topNodes: Table of the top differential nodes
In markrobinsonuzh/treeAGG: Tree Aggregation

Description Usage Arguments Value Author(s) Examples

topNodes extracts the top differential nodes in a data frame or a list of data frame for a given pair of group, ranked by p-value or absolute change (e.g. log-fold change)

1 2	topNodes(data, sort.by = "FDR", decreasing = FALSE, col.rowData = NULL, col.linkData = NULL, use.assays = NULL)

`data`	A tree summarizedExperiment object output from `treeAGG`
`sort.by`	The name of the column to sort by
`decreasing`	A logical value. TRUE or FALSE. Should the sorting be decreasing or not.
`col.rowData`	The names of columns to be extracted from `rowData`.
`col.linkData`	The names of columns to be extracted from `linkData`.
`use.assays`	A numeric vector. It specifies the result of which table should be shown. The default is NULL, all available results are shown; otherwise, the result of the specified table (via the table number) is shown. To recall, the table number is the number of table in `assays` (e.g. 1 represents the first table). If users forgot which tables have been used to do data analysis, they could check `use.assays` in the `metadata` of input `data`.

a list of data frame

Ruizhu HUANG

library(S4Vectors)
set.seed(1)
y <- matrix(rnbinom(300,size=1,mu=10),nrow=10)
colnames(y) <- paste(rep(LETTERS[1:3], each = 10), rep(1:10,3), sep = "_")
rownames(y) <- tinyTree$tip.label

rowInf <- DataFrame(nodeLab = rownames(y),
                    var1 = sample(letters[1:3], 10, replace = TRUE),
                    var2 = sample(c(TRUE, FALSE), 10, replace = TRUE))
colInf <- DataFrame(gg = factor(sample(1:3, 30, replace = TRUE)),
                    group = rep(LETTERS[1:3], each = 10))
toy_lse <- leafSummarizedExperiment(tree = tinyTree, rowData = rowInf,
                                    colData = colInf,
                                    assays = list(y, (2*y), 3*y))

toy_tse <- nodeValue(data = toy_lse, fun = sum, tree = tinyTree,
 message = TRUE)

# build the model
contrastList <- list(contrast1 = c(0, 0, 0, -1, 1),
                     contrast2 = c(0, -1, 1, 0, 0))
mod <- runEdgeR(obj = toy_tse, contrast = contrastList)
# results are stored as the column result_assay1, result_assay2, and
# result_assay3
(res <- rowData(mod, internal = TRUE))
# show results gained from the second element of the assasy
# sort by PValue
topNodes(mod, sort.by = "PValue", use.assays = 2)