get_back_res_meeting_min_occ: Create phyloseq and igraph objects of taxa meeting certain...
In ConesaLab/MDM: Complete MDM Network Analysis for 16S rRNA Data

Description Usage Arguments Value Examples

To create suitable networks for each environment, we model the most prevalent taxa, excluding rarer taxa that do not meet a strict sample prevalence threshold. Using the original OTU table information, we set a threshold of sample prevalence percentage and discard any taxa not meeting this threshold. All taxa that remain are saved to a new phyloseq object containing the updated, reduced OTU table information. The direct co-occurrence relationships are then measured for remaining taxa using the SpiecEasi MB neighborhood algorithm and stored in an igraph object.

1	get_back_res_meeting_min_occ(phylo, filter_val_percent = 0.4)

`phylo`	Phyloseq object of OTU table and taxonomic information
`filter_val_percent`	Minimum sample percent threshold (from 0 to 1) taxa must meet to be included in the environmental network. Default threshold is 0.4 (40 percent of all samples). Using this default threshold would mean that all taxa present in less than 40 percent of all samples will be discarded from network creation and subsequent steps.

Returns list of 2 objects.

`silva_phyloseqobj.final`	Phyloseq object including only taxa meeting sample prevalence threshold
`silva_spiec.graph`	Igraph object of direct co-occurrence relationships between these taxa calculated by SpiecEasi MB algorithm.

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

## The function is currently defined as
function (phylo, filter_val_percent = 0.4)
{
    silva_otus = phyloseq::otu_table(phylo)
    newsilva_taxa = phyloseq::tax_table(phylo)
    silva_filterobj = seqtime::filterTaxonMatrix(silva_otus,
        minocc = filter_val_percent * length(rownames(sample_data(phylo))),
        keepSum = TRUE, return.filtered.indices = TRUE)
    silva_otus.f = silva_filterobj$mat
    silva_taxa.f = newsilva_taxa[setdiff(1:nrow(newsilva_taxa),
        silva_filterobj$filtered.indices), ]
    dummyTaxonomy = c("D_0__dummy", "D_1__", "D_2__", "D_3__",
        "D_4__", "D_5__", "D_6__")
    silva_taxa.f = rbind(silva_taxa.f, dummyTaxonomy)
    rownames(silva_taxa.f)[nrow(silva_taxa.f)] = "0"
    rownames(silva_otus.f)[nrow(silva_otus.f)] = "0"
    silva_updatedotus = otu_table(silva_otus.f, taxa_are_rows = TRUE)
    silva_updatedtaxa = tax_table(silva_taxa.f)
    silva_phyloseqobj.final = phyloseq::phyloseq(silva_updatedotus,
        silva_updatedtaxa)
    silva_spiec.out = SpiecEasi::spiec.easi(silva_phyloseqobj.final,
        method = "mb", icov.select.params = list(rep.num = 20))
    silva_spiec.graph = igraph::adj2igraph(silva_spiec.out$refit,
        vertex.attr = list(name = taxa_names(silva_phyloseqobj.final)))
    phylo_and_graph <- list(silva_phyloseqobj.final, silva_spiec.graph)
    return(phylo_and_graph)
  }