Filtering feature who is low relative abundance or unclassified — filter

whether to filter the low relative abundance or unclassified feature by the threshold. Here, we choose the following criterion:

Feature more than Mean absolute or relative abundance across all samples;
Feature more than Minimum absolute or relative abundance at least one sample.

filter_abundance(
   object,
   level = c(NULL, "Kingdom", "Phylum", "Class",
           "Order", "Family", "Genus",
           "Species", "Strain", "unique"),
   cutoff_mean = c(100, 1e-04, 2),
   cutoff_one = c(1000, 1e-03, 3),
   unclass = TRUE)

Arguments

object: (Required). a phyloseq::phyloseq or SummarizedExperiment::SummarizedExperiment object.
level: (Optional). character. taxonomic level to summarize, default the top level rank of the ps. taxonomic level(Kingdom, Phylum, Class, Order, Family, Genus, Species, Strains; default: NULL).
cutoff_mean: (Optional). numeric. Threshold for Mean absolute (integer) or relative (float) abundance all samples (default, 0).
cutoff_one: (Optional). numeric. Threshold for Minimum absolute (integer) or relative (float) abundance at least one sample (default, 0).
unclass: (Optional). logical. whether to filter the unclassified taxa (default TRUE).

Value

a phyloseq::phyloseq or SummarizedExperiment::SummarizedExperiment object, where each row represents a feature and each col represents the feature abundance of each sample.

References

Thingholm, Louise B., et al. "Obese individuals with and without type 2 diabetes show different gut microbial functional capacity and composition." Cell host & microbe 26.2 (2019): 252-264.

Author

Created by Hua Zou (11/30/2021 Shenzhen China)

Examples


if (FALSE) {
# phyloseq object
 data("Zeybel_2022_gut")
 Zeybel_2022_gut_counts <- phyloseq::transform_sample_counts(
 Zeybel_2022_gut, function(x) {round(x * 10^7)})

 # absolute abundance
 ps <- filter_abundance(
   object = Zeybel_2022_gut_counts,
   level = NULL,
   cutoff_mean = 100,
   cutoff_one = 1000,
   unclass = FALSE)

 # relative abundance
 ps <- filter_abundance(
   object = Zeybel_2022_gut,
   level = "Phylum",
   cutoff_mean = 1e-04,
   cutoff_one = 1e-03,
   unclass = TRUE)


# SummarizedExperiment object
data("Zeybel_2022_protein")
filter_abundance(
  object = Zeybel_2022_protein,
  cutoff_mean = 5,
  cutoff_one = 8)
}