Skip to contents

Derive starter sample metadata from count sample names

Source: R/input_preparation.R
derive_vista_metadata.Rd

derive_vista_metadata() creates a starter sample_info table from count sample names. It is intended for projects where users have count columns but do not yet have a separate metadata sheet. The derived table can be edited, passed through read_vista_metadata(), and then aligned with match_vista_inputs().

Usage

derive_vista_metadata(
  counts,
  column_geneid = NULL,
  sample_names = NULL,
  parser = c("auto", "split", "regex", "none"),
  split = "_",
  fields = NULL,
  pattern = NULL,
  sample_column = "sample_names",
  repair_sample_names = c("auto", "none"),
  return_type = c("data.frame", "template"),
  verbose = TRUE
)

Arguments

counts

Count input accepted by read_vista_counts(), or the list returned by read_vista_counts().

column_geneid

Optional gene identifier column for raw tabular count inputs. Ignored when counts is the list output of read_vista_counts().

sample_names

Optional explicit sample names to derive metadata from. When supplied, these override names extracted from counts.

parser

Metadata parsing mode. "auto" tries a simple delimiter-based split when sample names have a consistent structure. "split" uses split explicitly. "regex" uses pattern. "none" returns only the sample_names column.

split

Delimiter used when parser = "split" or when "auto" chooses split-based parsing.

fields

Optional field names for parsed metadata columns. When omitted, VISTA uses part_1, part_2, etc.

pattern

Regular expression used when parser = "regex". Capture groups are mapped to fields in order.

sample_column

Name of the sample identifier column in the returned metadata. Default is "sample_names".

repair_sample_names

Strategy passed to read_vista_counts() when sample names are taken from counts. One of "auto" or "none".

return_type

Return "data.frame" (default) or "template". Both return a data frame; "template" adds empty placeholder columns for group and batch.

verbose

Logical; print an informational derivation summary.

Value

A data frame containing sample_names plus any parsed metadata columns.

Examples

data("count_data", package = "VISTA")
data("sample_metadata", package = "VISTA")

counts_in <- count_data[seq_len(8), c("gene_id", sample_metadata$sample_names[seq_len(6)]), drop = FALSE]
meta <- derive_vista_metadata(
  counts_in,
  column_geneid = "gene_id",
  parser = "regex",
  pattern = "SRR(\\d+)",
  fields = "run_id"
)
#> Derived metadata for 6 samples using parser "regex".
head(meta)
#>   sample_names  run_id
#> 1   SRR1039508 1039508
#> 2   SRR1039509 1039509
#> 3   SRR1039512 1039512
#> 4   SRR1039513 1039513
#> 5   SRR1039516 1039516
#> 6   SRR1039517 1039517