Description Usage Arguments Value References Examples
View source: R/bottom_up_inference.R
This function implements the (one-stage and two-stage) bottom-up approach to testing hypotheses that have a branching tree dependence structure, with false discovery rate control. Our motivating example comes from testing the association between a trait of interest and groups of microbes that have been organized into operational taxonomic units (OTUs) or amplicon sequence variants (ASVs). Given p-values from association tests for each individual OTU or ASV, we would like to know if we can declare that a certain species, genus, or higher taxonomic grouping can be considered to be associated with the trait. If a large proportion of species from that genus influence the trait, we should conclude the genus influences the trait. Conversely if only a few of the species from a genus are non-null, then a better description of the microbes that influence occurrence of the trait is a list of associated species. Finding taxa that can be said to influence a trait in this sense is the first goal of our approach. The second goal is to locate the highest taxa in the tree for which we can conclude many taxa below, but not any ancestors above, influence risk; we refer to such taxa as driver taxa.
1 2 |
anno.table |
an |
pvalue.leaves |
a vector of p-values at leaf nodes, the names of which
should match the values in the last column of |
na.symbol |
a character string which is to be inpretreted as |
far |
nominal false assignment rate (FAR), the error rate in analogy
with the false discovery rate. If |
tau |
a pre-specified constant to prevent nodes with large p-values from
being detected if a large number (say, |
is.weighted |
a logical value indicating whether the weighted or unweighted test is performed. The default is ‘TRUE’. |
A list consisting of
results.by.level |
a data frame that gives the number of detected nodes at each level along with information on the test for that level. |
results.by.node |
a data frame that gives detailed results at each node (i.e., leaf and inner nodes), including the (derived) p-value, the indicator of a detection or not, and the indicator of a driver node or not. |
tree |
A tree structure used in the function |
Li, Y., Satten, GA., Hu Y.-J. "A bottom-up approach to testing hypotheses that have a branching tree dependence structure, with false discovery rate control" XXX(2018)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 | data(IBD)
## one-stage weighted bottom-up test on the IBD data
test.1 = bouth(anno.table = IBD$tax.table, pvalue.leaves = IBD$pvalue.otus,
na.symbol = "unknown", far = 0.1, is.weighted = TRUE)
## extract all detected nodes
test.1$results.by.node[test.1$results.by.node$is.detected, ]
## extract all detected driver nodes
test.1$results.by.node[test.1$results.by.node$is.driver,]
## two-stage (weighted) bottom-up test on the IBD data
test.2 = bouth(anno.table = IBD$tax.table, pvalue.leaves = IBD$pvalue.otus,
na.symbol = "unknown", far = c(0.05, 0.05))
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.