Filter Phyloseq Object

Filter Phyloseq Object

filterSampleData(x, ...)

filterTaxaData(x, ...)

filterTaxaByNames(x, ids = NULL, keep = TRUE)

filterSampleByNames(x, ids = NULL, keep = TRUE)

Arguments

x

phyloseq object

...

Option to pass on to filter. Only when using filterSampleData or filterTaxaData sample data or taxonomy table as tibble.

ids

A list of sample ids or taxa ids to filter. Only when using filterSampleByNames or filterTaxaByNames

keep

Logical. Default is TRUE. Only when using filterSampleByNames or filterTaxaByNames

Value

Either filtered phyloseq

Details

These are alternative to subset and prune functions in phyloseq.

The filterSampleData does subsetting similar to phyloseq::subset_samples. The filterTaxaData does subsetting similar to phyloseq::subset_taxa.

The filterSampleByNames does pruning similar to phyloseq::prune_samples. The filterTaxaByNames does pruning similar to phyloseq::prune_taxa.

The naming is using filter because it uses the dplyr::filter function.

Note: filterSampleData will additionally remove any taxa that are zero across all samples i.e. not present is any remaining samples after filtering. filterTaxaData will additionally remove samples that do not have any of the selected taxa, i.e. samples that sum to zero. filterSampleByNames will additionally remove any taxa that are not present is any remaining samples after filtering. filterTaxaByNames will additionally remove samples that do not have any of the selected taxa, i.e. samples that sum to zero.

References

Shetty SA (2020). Utilities for microbiome analytics. https://github.com/RIVM-IIV-Microbiome/biomeUtils

Author

Sudarshan A. Shetty

Examples

library(biomeUtils)
data("FuentesIliGutData")

# Filter from tables subset_*
ps.filtered.samples  <- filterSampleData(FuentesIliGutData,
                                         ILI == "C" & BMI < 26)

ps.filtered.samples
#> phyloseq-class experiment-level object
#> otu_table()   OTU Table:         [ 901 taxa and 91 samples ]
#> sample_data() Sample Data:       [ 91 samples by 61 sample variables ]
#> tax_table()   Taxonomy Table:    [ 901 taxa by 7 taxonomic ranks ]
#> phy_tree()    Phylogenetic Tree: [ 901 tips and 900 internal nodes ]

ps.filtered.taxa  <- filterTaxaData(FuentesIliGutData,
                                    Phylum=="Firmicutes")

ps.filtered.taxa
#> phyloseq-class experiment-level object
#> otu_table()   OTU Table:         [ 758 taxa and 589 samples ]
#> sample_data() Sample Data:       [ 589 samples by 61 sample variables ]
#> tax_table()   Taxonomy Table:    [ 758 taxa by 7 taxonomic ranks ]
#> phy_tree()    Phylogenetic Tree: [ 758 tips and 757 internal nodes ]

# Filter by names like prune_*
sams.select <- sample_names(FuentesIliGutData)[1:10]
ps.filter.by.sam.names <- filterSampleByNames(FuentesIliGutData,
                                              ids = sams.select,
                                              keep = TRUE)

tax.select <- taxa_names(FuentesIliGutData)[1:10]
ps.filter.by.tax.names <- filterTaxaByNames(FuentesIliGutData,
                                            ids = tax.select,
                                            keep = TRUE)