Return the soma or root position of MANC bodyids

manc_rootpos returns the root point of all neurons and is a convenience wrapper around manc_somapos.

Usage

manc_somapos(
  ids = NULL,
  details = FALSE,
  duplicates = !details,
  somatags = c("soma", "tosoma", "root", "all"),
  node = "neutu",
  cache = TRUE,
  clio = TRUE
)

manc_rootpos(
  ids = NULL,
  details = FALSE,
  duplicates = !details,
  node = "neutu",
  cache = TRUE,
  clio = TRUE
)

Arguments

ids: MANC bodyids in any form compatible with manc_ids. The default ids=NULL returns the position of all known soma points.
details: Whether to include details such as bodyids in the return.
duplicates: Whether to include cases where the same bodyid has multiple soma annotations. Default is to include duplicates when details=TRUE.
somatags: Which annotations to include. The default is to include both "soma" and "tosoma" tags (see details). "all" is an alias for c("soma", "tosoma", "root").
node: The DVID node to use for XYZ to bodyid lookups. See manc_xyz2bodyid.
cache: whether to add a 5 min cache for the position lookups.
clio: Whether to include point annotations from clio as well as the contents of mancsomapos.

Value

When details=FALSE an Nx3 matrix. When TRUE, a data.frame containing XYZ position, bodyid, source and tag

Details

Currently there are three sources of soma information, the mancsomapos data.frame distributed with the malevnc package, a key value annotation stored in DVID and and point annotations stored in Clio. Currently DVID is considered the main repository of information with updates to clio.

The mancsomapos data.frame was the result of initial annotation work principally by Chris Ordish at FlyEM who marked two points to define a sphere. It contains some cases where the midpoint between the two marked points does not land on the intended soma; there are also cases where the segmentation inside the soma is not contiguous e.g. because the nucleus is a separate body. These annotations were reviewed by FlyEM in June 2021 (see e.g. Slack June 3rd and Slack June 10th).

Clio point annotations were added after predicting a set of objects without an annotated soma that might be intrinsic VNC neurons based on features such as pre- and post-synapse numbers, size, skeleton path length etc. These point annotation include "tosoma" positions not represented in DVID.

soma tags are placed only on somata visible within the dataset
tosoma tags are placed on cell body fibres leading to the soma of an intrinsic neuron that may be missing (e.g. because of damage to the specimen or because of difficulty following the cell body fibre all the way to the soma due to segmentation issues.)
root tags are placed on a part of the neuron that is likely to be the closest part of the neuron to the soma within the image volume. Two example use cases are DNs or sensory neurons whose soma is outside the volume.

Examples