Skip to main content

IMAP: Integrated Microbiome Analysis Pipelines: End-to-End Practical User Guides Using Integrated Approaches

4 Creating phyloseq Objects

In this section, we demonstrate how to create phyloseq objects to organize and structure microbiome data for analysis. These objects serve as the foundation for downstream analyses and visualization.

4.1 From mothur output 

library(tidyverse)
library(phyloseq)
library(microbiome)
library(ape)

metadata <- read_csv("data/mothur/mothur_tidy_metadata.csv",show_col_types = FALSE) %>% 
  tibble::column_to_rownames("group") %>% 
  sample_data(metadata)

otutable <- read_csv("data/mothur/mothur_tidy_otutable.csv",show_col_types = FALSE) %>% 
  pivot_wider(id_cols = group, names_from = OTU, values_from = count) %>% 
  tibble::column_to_rownames("group") %>% 
  otu_table(otutable, taxa_are_rows = FALSE)

taxonomy <- read_csv("data/mothur/mothur_tidy_taxonomy.csv", show_col_types = FALSE) %>%
  tibble::column_to_rownames("OTU") %>%
  as.matrix() %>% 
  tax_table(taxonomy)

mt_ps_raw_basic <- merge_phyloseq(metadata, otutable, taxonomy)

library(ape)
mt_ps_tree = rtree(ntaxa(mt_ps_raw_basic), rooted=TRUE, tip.label=taxa_names(mt_ps_raw_basic))

mt_ps_raw <- phyloseq::merge_phyloseq(mt_ps_raw_basic, mt_ps_tree)

mt_ps_rel <- phyloseq::transform_sample_counts(mt_ps_raw, function(x){x / sum(x)})

mt_ps_df_raw <- psmelt(mt_ps_raw)

mt_ps_df_rel <- psmelt(mt_ps_raw)

save(mt_ps_tree, mt_ps_raw, mt_ps_rel, mt_ps_df_raw, mt_ps_df_rel, file = "data/mothur_phyloseq_objects.rda")

4.1.1 Review the Mothur phyloseq object 

mt_ps_raw
phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 920 taxa and 10 samples ]
sample_data() Sample Data:       [ 10 samples by 2 sample variables ]
tax_table()   Taxonomy Table:    [ 920 taxa by 6 taxonomic ranks ]
phy_tree()    Phylogenetic Tree: [ 920 tips and 919 internal nodes ]

4.2 From QIIME2 output 

library(tidyverse)
library(phyloseq)
library(microbiome)
library(ape)

metadata <- read_csv("data/qiime2/qiime2_tidy_metadata.csv",show_col_types = FALSE) %>% 
  tibble::column_to_rownames("sample_id") %>% 
  sample_data(metadata)

otutable <- read_csv("data/qiime2/qiime2_tidy_otutable.csv",show_col_types = FALSE) %>% 
  pivot_wider(id_cols = sample_id, names_from = feature, values_from = count) %>% 
  tibble::column_to_rownames("sample_id") %>% 
  otu_table(otutable, taxa_are_rows = FALSE)

taxonomy <- read_csv("data/qiime2/qiime2_tidy_taxonomy.csv", show_col_types = FALSE) %>%
  tibble::column_to_rownames("feature") %>%
  as.matrix() %>% 
  tax_table(taxonomy)

q2_ps_raw_basic <- merge_phyloseq(metadata, otutable, taxonomy)

library(ape)
q2_ps_tree = rtree(ntaxa(q2_ps_raw_basic), rooted=TRUE, tip.label=taxa_names(q2_ps_raw_basic))

q2_ps_raw <- phyloseq::merge_phyloseq(q2_ps_raw_basic, q2_ps_tree)

q2_ps_rel <- phyloseq::transform_sample_counts(q2_ps_raw, function(x){x / sum(x)})

q2_ps_df_raw <- psmelt(q2_ps_raw)

q2_ps_df_rel <- psmelt(q2_ps_raw)

save(q2_ps_tree, q2_ps_raw, q2_ps_rel, q2_ps_df_raw, q2_ps_df_rel, file = "data/qiime2_phyloseq_objects.rda")

4.3 Review the QIIME2 phyloseq object 

q2_ps_raw
phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 95 taxa and 128 samples ]
sample_data() Sample Data:       [ 128 samples by 12 sample variables ]
tax_table()   Taxonomy Table:    [ 95 taxa by 6 taxonomic ranks ]
phy_tree()    Phylogenetic Tree: [ 95 tips and 94 internal nodes ]