Mirror an object in FAFB space using FAFB-JRC2018F registration

mirror_fafb(x, sample = NULL, subset = NULL, via = NULL, ...)

Arguments

x	An object containing 3D vertices (e.g. a neuron, surface, points)
sample	The starting brain space in which `x` lives - `FAFB14` by default if it is not otherwise other specified.
subset	Character, numeric or logical vector specifying on which subset of `X` the function `FUN` should be applied. Elements outside the subset are passed through unmodified.
via	(optional, for expert use only) Intermediate template brain that the registration sequence must pass through. See details and `xform_brain` for more information.
...	Additional arguments passed to `xform_brain` and `mirror_brain`

Value

A mirrored version of the neuron/surface or other 3D object.

Details

This function works by mapping the input data to JRC2018F, a high quality symmetric female brain template. The data are then flipped (since JRC2018F is symmetric no other action is required) and then returned to the starting space. In order to ensure that the Bogovic et al bridging registrations are preferred the via argument will be filled with the value "FAFB14um" i.e. FAFB14 calibrated in microns. This works because the Bogovic registrations were based on a mock synaptic staining using synapse locations that had been rescaled to microns before calculating a registration with ANTs to JRC2018F.

Examples

# \donttest{
# transform arbitrary location (in nm)
mirror_fafb(cbind(388112, 162988, 132840))
#>          [,1]     [,2]     [,3]
#> [1,] 667366.9 174034.2 102780.7
# same but in units of microns
mirror_fafb(cbind(388.112, 162.988, 132.84), sample='FAFB14um')
#>          [,1]     [,2]     [,3]
#> [1,] 667.3669 174.0342 102.7807
# no messages
mirror_fafb(cbind(388112, 162988, 132840), Verbose=FALSE)
#>          [,1]     [,2]     [,3]
#> [1,] 667366.9 174034.2 102780.7

# transform arbitrary location (from and to raw pixel coordinates)
voxel.size=c(4,4,40)
mirror_fafb(cbind(97028, 40747, 3321)* voxel.size)/c(voxel.size)
#>          [,1]     [,2]     [,3]
#> [1,] 166841.7 43508.56 2569.518

library(nat.flybrains)
if(require('elmr', quietly = TRUE)) {
FAFB.surf.m <- mirror_fafb(FAFB14.surf)
wire3d(as.mesh3d(FAFB.surf), col='blue')
wire3d(as.mesh3d(FAFB.surf.m), col='red')
}

if (FALSE) {
# you can also use with points in https://flywire.ai space
# converts from raw (voxel) coords used by Neuroglancer to nm and back again
voxel.size=c(4,4,40)
mirror_fafb(cbind(120508, 46292, 1667)*voxel.size, sample='FlyWire')/voxel.size
# compare with just using FAFB14
mirror_fafb(cbind(120508, 46292, 1667)*voxel.size)/voxel.size
}
# }