A SPIEC-EASI pipeline for inferring a sparse inverse covariance matrix within and between multiple compositional datasets, under joint sparsity penalty.
multi.spiec.easi(datalist, method = "glasso", sel.criterion = "stars",
verbose = TRUE, pulsar.select = TRUE, pulsar.params = list(), ...)# S3 method for list
spiec.easi(data, ...)
list of non-normalized count OTU/data tables (stored in a matrix, data.frame or phyloseq/otu_table) with samples on rows and features/OTUs in columns
estimation method to use as a character string. Currently either 'glasso' or 'mb' (meinshausen-buhlmann's neighborhood selection)
character string specifying criterion/method for model selection. Accepts 'stars' and 'bstars' [default]
flag to show progress messages
flag to perform model selection. Choices are TRUE/FALSE/'batch'
list of further arguments to pulsar or batch.pulsar. See the documentation for pulsar.params.
further arguments to sparseiCov / huge
non-normalized count OTU/data table with samples on rows and features/OTUs in columns. Can also be list of phyloseq objects.
Can also run spiec.easi on a list and S3 will dispatch the proper function.
spiec.easi