Return metadata about CRANTb neurons from the project seatable

crant_meta() is a thin wrapper around fafbseg::cam_meta() that points it at the CRANTb seatable instance (cloud.seatable.io, table CRANTb_meta). By default it returns the underlying seatable column names unchanged. Set normalise_colnames = TRUE to translate them to the coconat-friendly schema (id, class, type, etc.), which also enables query aliases such as "/class:descending" or "/type:NSC". Plain root id vectors are also accepted.

Usage

crant_meta(
  ids = NULL,
  ignore.case = FALSE,
  fixed = FALSE,
  version = NULL,
  timestamp = NULL,
  unique = FALSE,
  normalise_colnames = FALSE,
  translate_ids = NA,
  drop_status = "DUPLICATED",
  ...
)

Arguments

ids: Root ids (character/int64 vector) or a single query string such as "/class:descending" (see fafbseg::cam_meta()). NULL returns the full table.
ignore.case: For query strings, whether to ignore case.
fixed: For query strings, whether to treat queries as fixed strings rather than regular expressions.
version, timestamp: Optional CAVE materialisation version / timestamp to which root ids in the returned table should be mapped.
unique: Whether to drop duplicate id rows.
normalise_colnames: Logical; if TRUE, rename metadata columns to the coconat/crantr schema (root_id -> id, super_class -> class, etc.) and translate query aliases such as class: -> super_class: and bare tokens like "NSC" -> "cell_type:NSC".
translate_ids: Whether to bring explicitly supplied ids forward to the requested version/timestamp before matching, so stale ids don't silently drop out of the result (see fafbseg::cam_meta() for details). NA (the default) enables this automatically whenever version or timestamp is supplied.
drop_status: Character vector of status tokens whose rows are dropped before any query/join/unique step. Defaults to CRANT's capitalised multi-select duplicate marker "DUPLICATED"; pass NULL to keep known duplicates. See fafbseg::cam_meta().
...: Passed to fafbseg::cam_meta() (and via that to fafbseg::flytable_cached_table()).

Value

By default, a data.frame with the underlying seatable columns such as root_id, super_class, cell_class, cell_type, cell_subtype, cell_instance, hemilineage, side, nerve, tract. When normalise_colnames = TRUE, these are renamed to the crantr/coconat schema (id, class, subclass, type, subtype, instance, lineage) and side is normalised to "L"/"R".

Details

Disk caching with delta sync is handled by fafbseg::flytable_cached_table() under the hood, so repeated calls are fast. Pass expiry, refresh etc. via ... to control cache behaviour.

Internally crant_meta() uses withr::local_options() to point fafbseg's flytable client at the CRANTb seatable and withr::local_envvar() to temporarily pass CRANTTABLE_TOKEN through as FLYTABLE_TOKEN before calling fafbseg::cam_meta(). Requires CRANTTABLE_TOKEN to be set; see crant_table_set_token().

Examples

if (FALSE) { # \dontrun{
# all descending neurons
crant_meta("/super_class:descending")
# a specific descending neuron type
crant_meta("/cell_type:MDN")
# regex queries match anywhere in the field, e.g. all central complex
# (CX) neuron classes
crant_meta("/super_class:CX.+")

# the coconat-friendly schema and its query aliases
crant_meta("/class:descending", normalise_colnames = TRUE)
crant_meta("/type:MDN", normalise_colnames = TRUE)
# a bare token is treated as a cell_type/type query
crant_meta("ExR1", normalise_colnames = TRUE)
# ... equivalently, on the raw seatable schema
crant_meta("/cell_class:olfactory_projection_neuron")

crant_meta(c("576460752684030043", "576460752653449509"))
} # }