ScientificOSX

These are the steps I took recently to setup a new MacbookPro (Sonoma OS 14.7.2) for scientific/bioinformatics use.

Step 1, base software:

Log in to App Store with login. Xcode and other needed items are autodownloaded by homebrew via the App Store.

Start Xcode & accept terms.

Install homebrew, update

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install.sh)"

brew analytics off

brew update && brew upgrade

brew config

Extras

#skip this, its pedantic

#update bash
bash --version

brew install bash

#The newer version of bash is located at /usr/local/bin/bash:
ls -l /usr/local/bin/bash
cat /etc/shells


# If not in /etc/shells, append it:
sudo -i
echo /usr/local/bin/bash >> /etc/shells

#Set default shell to /usr/local/bin/bash, run the following chsh command
chsh -s /usr/local/bin/bash

#should have changed to latests
bash --version

Create

touch ~/.bash_profile
# to edit use nano 
nano ~/.bash_profile

Include this in ~/.bash_profile & save.

export PATH=/usr/local/bin:/usr/local/sbin:$PATH
alias rc="source ~/.bash_profile"

#alias man="_() { echo $1; man -M $(brew --prefix)/opt/coreutils/libexec/gnuman $1 1>/dev/null 2>&1;  if [ "$?" -eq 0 ]; then man -M $(brew --prefix)/opt/coreutils/libexec/gnuman $1; else man $1; fi }; _"

source ~/.bash_profile

Xquartz

brew install --cask xquartz

JAVA

brew tap AdoptOpenJDK/openjdk
brew install --cask adoptopenjdk11

R and R studio:

install https://rud.is/rswitch/ Install latest R: https://cran.r-project.org/bin/macosx/ RStudio: https://www.rstudio.com/products/rstudio/download/#download

At this time, I need 4.2 and 4.4 so I installed them both respectively.

Multiple versions of R

Blindly using the binary installer (*.pkg) version available through CRAN will remove the current version of the r framework (including your package libraries).

Thus, prior to installing the new version of R is trick OSX into forgetting that R is already installed on your system by using rswitch.

https://rud.is/rswitch/guide/index.html

#this is old stuff hoarding just in case

pkgutil --pkgs #shows what is installed
#updated for catalina, and I installed 4.0 first then 3.6 and the below code worked great.
#Make sure you copy over the RSwitch.App to your applications folder and run it.

sudo pkgutil --forget org.R-project.R.fw.pkg
sudo pkgutil --forget org.r-project.x86_64.tcltk
sudo pkgutil --forget org.R-project.R.GUI.pkg
             
             

Base Needs

brew install hdf5 git freetype
brew install cmake automake fontconfig jpeg libtiff libpng libffi doxygen
brew install glib libcaca libconfig libtasn1 libtool libunistring zlib icu4c
brew install tmux

Python and Libs

management of different python versions

There are a number of tools available that enable the installation and management of different python versions.

pyenv is one tool to do this. One could also use anaconda or miniconda or similar.

#skip for now, later we can use a flavor of conda
brew install pyenv
echo "eval "$(pyenv init -)"" >> ~/.bash_profile
source ~/.bash_profile
pyenv install --list

#see list for latest
#pyenv install 3.9.1 # 3.9 does not support Numpy
pyenv install  3.8.7
pyenv install 2.7.18

#set globally which python
pyenv global 3.8.7
pyenv version

#update pip
pip install --upgrade pip

virtual python environments

To have different python projects with different dependencies and versions of python; environments help isolate them.

to create python virtual environments using pipenv:

#skip for now, later we can use a flavor of conda

brew install pipenv


mkdir MyPyTest
cd MyPyTest
pipenv --python 3.8

#to install packages use pipenv within the environment

pipenv install ipykernel

Note: Jupyter Notebooks are able to work with virtual environments so that you are able to run the notebooks for a project in the correct project environment.

#skip for now, later we can use a flavor of conda

#change myenv to a name
python3 -m ipykernel install --user --name myenv --display-name "Python (myenv)"

Python Packages

jupyter notebook can be invoked which will automatically open a window in your browser.

pip3 install notebook
pip3 install h5py
pip3 install umap-learn

Extra installs

#skip for now
Install latest gfortran https://github.com/fxcoudert/gfortran-for-macOS/releases

R packages

Note if you have multiple R environments, you may want to do this for each.

#update Rprofile by adding the line below and saving
nano ~/.Rprofile.site
# local({options(repos = BiocManager::repositories())})

install.packages(c("devtools", "BiocManager", "remotes"), dependencies=TRUE, ask = FALSE)

BiocManager::install(c("BiocParallel", "DropletUtils", "ExperimentHub", "SingleCellExperiment"), dependencies=TRUE)
BiocManager::install(c("SingleR", "celldex", "DESeq2", "MAST", "scDblFinder", "ComplexHeatmap", "clusterProfiler", "RUVSeq", "escape", "GSVA", "decoupleR", "OmnipathR", "monocle", "princurve", "vegan", "RSelenium", "wordcloud" , "ggpubr"), dependencies=TRUE)
BiocManager::install(c("org.Hs.eg.db", "org.Mm.eg.db", "HSMMSingleCell", "DelayedMatrixStats", "DESeq2", "genefilter", "preprocessCore", "demuxmix"), dependencies=TRUE)
BiocManager::install(c("preprocessCore", "demuxmix"), dependencies=TRUE)

devtools::install_github("karthik/wesanderson")
devtools::install_github("marchinilab/SDAtools")
devtools::install_github("davidsjoberg/ggsankey")
devtools::install_github("VPetukhov/ggrastr")
remotes::install_github("juba/rmdformats")

devtools::install_github(repo = "bimberlabinternal/cellmembrane", dependencies = TRUE)
devtools::install_github(repo = "eisascience/scCustFx", dependencies = TRUE)
devtools::install_github(repo = "bimberlab/cellhashR", ref = "master", dependencies = TRUE)
devtools::install_github(repo = "bimberlabinternal/Rdiscvr", dependencies = TRUE)
devtools::install_github(repo = "bimberlabinternal/rira", dependencies = TRUE)

# install.packages("XML", repos = "http://www.omegahat.net/R")

cask software

brew install --cask cyberduck
brew install --cask docker