Export a phyloseq object to GBIF MDT template CSV/TSV files

Write the OTU table, the taxonomy table and the sample data of a phyloseq object to separate delimited text files, one per table, following the GBIF Metabarcoding Data Toolkit (MDT) template layout. This is the plain-text (CSV/TSV) counterpart of phyloseq_to_MDT_excel(), which writes a single multi-sheet .xlsx file.

The MDT template expects the OTU table with OTU IDs in rows and sample IDs in columns (sequence read counts in the cells), and the sample and taxonomy tables keyed by a leading id column. The reference sequences, when available, are appended as a DNA_sequence column of the taxonomy file (Darwin Core DNA-derived-data extension term).

When check_dwc = TRUE (the default), a lightweight Darwin Core compliance check warns about recommended sample-level terms that are missing from sample_data (decimalLatitude, decimalLongitude, eventDate). This is a non-blocking helper, not a full validation of the GBIF MDT template.

Usage

phyloseq_to_MDT_csv(
  physeq,
  path = ".",
  prefix = "",
  sep = ",",
  check_dwc = TRUE
)

Arguments

physeq

(required) a phyloseq-class object.

path

(character, default ".") Directory where the files are written. Created (recursively) if it does not exist.

prefix

(character, default "") Optional prefix prepended to each output file name (e.g. "data_fungi_").

sep

(character, default ",") Field separator. Use "\t" to write tab-separated (TSV) files, the format favored by the GBIF MDT validator. The file extension follows sep (.csv for ",", .tsv otherwise).

check_dwc

(logical, default TRUE) If TRUE, warn about recommended Darwin Core sample-level terms missing from sample_data.

Value

Invisibly returns a named character vector of the written file paths (OTU_table, Taxonomy, Samples).

Details

See the GBIF Metabarcoding Data Toolkit for the expected input format. The MDT accepts both TSV and XLSX uploads.

See also

phyloseq_to_MDT_excel()

Author

Adrien Taudière

Examples

# \donttest{
out_dir <- file.path(tempdir(), "mdt_csv")
files <- phyloseq_to_MDT_csv(clean_pq(data_fungi_mini), path = out_dir)
#> Warning: ! Recommended Darwin Core sample terms missing from sample_data:
#>   "decimalLatitude", "decimalLongitude", and "eventDate".
#> ℹ Add them before GBIF MDT submission if available.
#> MDT template files written to /tmp/RtmpMeUGPY/mdt_csv
file.exists(files)
#> [1] TRUE TRUE TRUE
unlink(out_dir, recursive = TRUE)
# }
if (FALSE) { # \dontrun{
# Tab-separated output (MDT-favored TSV)
phyloseq_to_MDT_csv(data_fungi_mini, path = "mdt", sep = "\t")
} # }