Linux installation

Step by step guide

0. Installation should take less than 1 h. Most time will be spent compiling R and python packages. Most R packages will have to be compiled on Linux systems.

1. Install library dependencies. We have tested this on Ubuntu 24.04.1 LTS. The package list might therefore vary depending on your distribution.

Install dev libraries

sudo apt-get install libcurl4-openssl-dev libxml2-dev cmake libgdal-dev libmagick++-dev libudunits2-dev libharfbuzz-dev libfribidi-dev

2. Install R base for Ubuntu and RStudio. We tested this on R version 4.4.1. If you use another version of R the package dependencies might not be resolved during renv::init().

3. Start RStudio and install renv.

Install renv

install.packages("renv")

4. Install package dependencies. Download the renv.lock file and create the R-environment. You must set the current working directory to the directory where the renv.lock file is located. Select “1” to restore the project from the lockfile.

Tip

Unix systems have three main signs to specify directories

Common path directories

~ defines the home directory
. defines the current directory
.. defines the parent directory

~/Documents is shortform for /home/dom/Documents

Init R-environment

# An example would be
# setwd("~/Cecelia")
setwd("PATH/TO/RENV/LOCK/FILE")
renv::init()

5. Load the environment you have just created and install Cecelia package.

Install Cecelia package

renv::load()
renv::install("schienstockd/cecelia")

6. You must define a base directory where configuration files, models and the shiny app will be stored.

Define base directory

library(cecelia)
cciaSetup("/REPLACE/WITH/YOUR/PATH")

In case you get stuck at any point and have to restart R, you need to redefine the path you are working on

Restart Cecelia

Sys.setenv(KMP_DUPLICATE_LIB_OK = "TRUE")
library(cecelia)
cciaUse("/REPLACE/WITH/YOUR/PATH")

5. Cecelia depends on a conda environment which must be created.

Install miniconda

reticulate::install_miniconda()

Create conda environment

cciaCondaCreate()

6. Download models for deep-learning segmentation, tracking and others.

Download models

cciaModels()

7. Create shiny app in base directory.

Create shiny app

cciaCreateApp()

8. Adjust config

You have to adjust the parameters in /YOUR/PATH/TO/cecelia/custom.yml to your system and download/install:

• bioformats2raw

Adjust config in text editor of RStudio

default:
  dirs:
    bioformats2raw: "/REPLACE/PATH/TO/bioformats2raw"
    projects: "/REPLACE/PATH/TO/project/directory/"
  volumes:
    home: "~/"
    computer: "/"
  python:
    conda:
      env: "r-cecelia-env"
      source:
        env: "r-cecelia-env"

9. Run the app.

Run Cecelia app

/PATH/TO/CECELIA/app/cecelia-Linux.sh