2024crant collects together R scripts and functions to work with the clonal raider ant brain (CRANTb) dataset. See https://github.com/flyconnectome/crantr for more details.
To use, install R + Rstudio and open the .Rproj file.
This repository includes an renv lockfile to manage dependencies. When you start R it should offer to install any packages that you do not have. Although this can be a nice idea in general, there is a specific reason in this case: Alex Bates' crantr package currently depends on a development branch of the fafbseg package with some wrinkles in the authorisation mechanisms.
Here is an R example of fishing upstream of a single Olfactory Projection Neuron (PN) to find its olfactory receptor input neurons (ORNs) and then going downstream from the ORNs to find the other PNs in the glomerulus. It seems that there are at least 3 in this case.
library(dplyr)
source("R/crant-funs.R")
crant_partner_summary()
# PN defining a glomerulus
# make sure id is up to date
g1=crantr::crant_latestid("576460752752912786")
# find all the inputs onto this neuron
g1.in=crant_partner_summary(g1, partners = 'i')
# copy/paste ids to neuroglancer, review them and pick out ORNs
g1.in %>% filter(weight>2) %>% with(clipr::write_clip(pre_id))
g1.orns <- fafbseg::flywire_ids('576460752719726635, 576460752719881771, 576460752720182571, 576460752736400683, 576460752736409899')
# Find ORN outputs,
g1.orns.o=crant_partner_summary(g1.orns, partners = 'o')
# review those, copy/paste ids to neuroglancer and then pick out additional PNs
g1.orns.o %>%
filter(weight>5) %>%
with(clipr::write_clip(post_id))