For the impatient ...
# install
if (!require("devtools")) install.packages("devtools")
devtools::install_github(c("jefferis/nat", "jefferislab/nat.nblast"))
# use
library(nat.nblast)
# run examples for search
example("nblast")
# run examples for clustering
example("nhclust")
# get overview help for package
?nat.nblast
# help for functions
?nblast
?nhclust
# run tests
library(testthat)
test_package("nat.nblast")
This R package implements the NBLAST neuron similarity algorithm described in a preprint available at http://dx.doi.org/10.1101/006346. In addition to basic pairwise comparison, the package implements search of databases of neurons. There is also suport for all x all comparison for a group of neurons. This can produce a distance matrix suitable for hierarchical clustering, which is also implemented in the package.
These tools are designed as an addon for the NeuroAnatomy Toolbox (nat) R package, which will be installed as dependency.
You will probably find the following online documentation helpful:
- http://jefferislab.github.io/nat.nblast - Online documentation for this R package
- http://jefferislab.org/si/nblast - Overview of NBLAST algorithm and online tools
This package has been released to CRAN (since v1.5), but we generally recommend installing the development version from GitHub, especially if you notice a bug.
install.packages("nat.nblast")
Use devtools to install the development version:
# install devtools if required
if (!require("devtools")) install.packages("devtools")
# then nat.nblast
devtools::install_github("nat.nblast", "jefferislab")
Note that this will also update the nat package to the latest development version from github. Windows users need Rtools to install this way.