Skip to contents

Set or append rowData annotations on a VISTA object

set_rowdata.Rd

Accepts a data.frame/tibble/DataFrame of gene-level annotations, aligns it to the VISTA row order, and stores it in rowData(x). Rows are matched by a key column (default: tries gene_id or rownames); Ensembl version suffixes can be stripped for matching.

Usage

set_rowdata(
  x,
  annotations = NULL,
  orgdb = NULL,
  key_col = NULL,
  keytype = NULL,
  columns = c("SYMBOL", "GENENAME", "ENSEMBL", "ENTREZID", "TXCHROM", "TXSTART", "TXEND"),
  drop_version = TRUE,
  overwrite = FALSE
)

Arguments

x

A VISTA object.

annotations

Optional data.frame/tibble/DataFrame with one row per gene and a column containing the gene IDs to match against rownames(x). If omitted, annotations are pulled from orgdb.

orgdb

Optional OrgDb object; when supplied (and annotations is NULL), annotations are retrieved via AnnotationDbi::select().

key_col

Name of the column in annotations that holds the gene IDs. If NULL, the function will try gene_id, gene, ENSEMBL, SYMBOL, or use rownames(annotations). Ignored when annotations is NULL and orgdb is used.

keytype

Key type for orgdb lookups (e.g., "ENSEMBL", "SYMBOL"). If NULL, inferred from rownames(x) (ENSEMBL if they start with "ENS", otherwise SYMBOL).

columns

Character vector of OrgDb columns to retrieve when using orgdb. Default: c("SYMBOL","GENENAME","ENSEMBL","ENTREZID","TXCHROM","TXSTART","TXEND"). The TXCHROM/TXSTART/TXEND fields carry basic genomic coordinates when available in the OrgDb.

drop_version

Logical; if TRUE, strips Ensembl version suffixes (e.g., .1) from both the VISTA rownames and the key column/keys before matching.

overwrite

Logical; if TRUE, replaces existing rowData. If FALSE, new columns are appended (overwriting by name when names collide).

Value

The updated VISTA object with rowData populated/appended.

Details

OrgDb packages rarely include full genomic coordinates; the default TXCHROM/TXSTART/TXEND columns may therefore be NA unless your OrgDb provides them. For reliable coordinates, fetch them from an EnsDb/TxDb (via genes() or biomaRt/AnnotationHub), build an annotation table keyed on your gene IDs, and supply that via the annotations argument. When fetching from an OrgDb, only columns available in that database will be filled.

Examples

if (FALSE) { # \dontrun{
# Load example VISTA object
data("count_data", package = "VISTA")
data("sample_metadata", package = "VISTA")

vista <- create_vista(
  counts = count_data[1:100, ],
  sample_info = sample_metadata[1:6, ],
  column_geneid = "gene_id",
  group_column = "cond_long",
  group_numerator = "treatment1",
  group_denominator = "control"
)

# Add annotations from OrgDb (human)
if (requireNamespace("org.Hs.eg.db", quietly = TRUE)) {
  vista <- set_rowdata(
    vista,
    orgdb = org.Hs.eg.db::org.Hs.eg.db,
    columns = c("SYMBOL", "GENENAME", "ENTREZID")
  )

  # View updated rowData
  head(rowData(vista))
}

# Or provide custom annotations
custom_annot <- data.frame(
  gene_id = rownames(vista)[1:10],
  custom_info = paste0("Info_", 1:10)
)
vista <- set_rowdata(vista, annotations = custom_annot, key_col = "gene_id")
} # }