README.md

In beanumber/cranium: Quantifying Radial Bridge Structures

An R package to quantify radial bridge images

Installation

cranium depends on XQuartz, which is not on CRAN. You can download from https://www.xquartz.org/. After install XQuartz, restart your R and continue cranium installation.

cranium depends on thehdf5r package for reading HDF5 files. You could download hdf5r from CRAN.

install.packages("hdf5r")

If you have trouble downloading hdf5r, the alternative is rhdf5 package. However, rhdf5 package is not available on CRAN--it is a bioconductor package. You must install it from Bioconductor. To install the package, you have to run the following code once:

install.packages("BiocInstaller",
                 repos = "http://bioconductor.org/packages/3.4/bioc")
source("https://bioconductor.org/biocLite.R")
biocLite("rhdf5")

If you are having trouble downloading rhdf5 using the code above, an alternative option would be downloading rhdf5 from bioconductor using the pak package: https://github.com/r-lib/pak. Also, check the most recent version of Bioconductor here: http://bioconductor.org/about/removed-packages/.

Note that you do not have to download rhdf5 if you successfuly downloaded hdf5r.

After installing rhdf5 or hdf5r, you can install cranium from GitHub using the following commands:

install.packages("devtools")
devtools::install_github("beanumber/cranium")

Library cranium package everytime you start your R session.

library(cranium)

Download sample data

To download sample data, You Too and WildType, you can run:

youtoo <- download_youtoo_data("folder") #where folder is the location you want to store the You Too data
wildtype <- download_wildtype_data("folder") #where folder is the location you want to store the Wild Type data
youtoo[[1]] #first you too data sample

Once the above is ran, the files are saved in a list and in the folder you set, each data sample can be indivudually picked from list by indexing.

Plot a 3D image of a brain

Now, you can simply point to a raw HDF5 file, and quickly render a 3D image of the data, along with our model of it.

file <- "~/Data/barresi/AT_1_Probabilities.h5"
library(cranium)
library(tidyverse)
tidy_brain <- file %>%
  read_h5() %>%
  tidy()
plot3d(tidy_brain)

Evaluate quadratic model fit

A quadratic model y = x^2+x is used to fit the data. You can evaluate the quadratic model fit on your sample. The default threshold for signal frequency is 0.9.

file%>%
  read_h5()%>%
  qmodel.brain()

If you are working with You-Too sample, you can change the data type and the default threshold for You-Too sample is 0.5. You can also adjust the threshold.

file%>%
  read_h5()%>%
  qmodel.brain(type="youtoo", threshold = 0.8)

beanumber/cranium documentation built on Oct. 17, 2019, 11:03 a.m.