Skip to content

Latest commit

 

History

History
 
 

02.integration

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
src
src
 
 
Makefile
Makefile
 
 
README.org
README.org
 
 

README.org

Integration analysis with Allen scRNA-seq data and snATAC-seq data

Allen scRNA-seq (10Xv3 + 10Xv2) and snRNA-seq (multiome)

  • Downloading: Under encoder, use aws s3 cp [allen data] [local_target]
  • They are saved in scanpy anndata format.
  • anndata.X is log-normalzied, which is CPM-log-normlization.
    • CPM: count per million

BICCN-CEMBA brain region

Installation of Seurat in tscc

conda config

  • conda config –add channels defaults
  • conda config –add channels bioconda
  • conda config –add channels conda-forge

conda install

  • install R using conda
    • conda create -n seurat r-base r-essentials

Install seurat4 after activating conda env

  • Install dependencies (having issues when installing them under R)
    • conda install -c conda-forge r-uwot
  • R: install.packages(“Seurat”)

Install Seurat5

  • Shell: conda install -c conda-forge r-uwot conda install -c anaconda hdf5 conda install -c conda-forge reticulate
  • R: install.packages(c(“R.utils”, “future”, “remotes”))
  • R: remotes::install_github(“bnprks/BPCells”)
  • R: BiocManager::install(“Rsamtools”) R: remotes::install_github(“stuart-lab/signac”, “seurat5”)
  • R: remotes::install_github(“satijalab/seurat”, “seurat5”) remotes::install_github(“satijalab/seurat-data”, “seurat5”) [OPTIONAL] remotes::install_github(“mojaveazure/seurat-disk”)
    • [MUST] Prepare to FastMNN, Harmony for integration
      • R: remotes::install_github(‘satijalab/seurat-wrappers’, ref = “seurat5”, force = TRUE, upgrade = “never”)
        • may need to usethis package to add GITHUB_PAT to rprofile
        • only this version has FastMNNIntegration
      • R: FastMNN BiocManager::install(“batchelor”)
      • R: Harmony install.packages(“harmony”)

Settings of Seurat v5

  • [MUST] R: options(Seurat.object.assay.version = ‘v5’) To use new Seurat v5 assays. Loading Seurat v5 beta version To maintain compatibility with previous workflows, new Seurat objects will use the previous object structure by default.
  • [MUST] R: options(future.globals.maxSize = 6e9) Even not use future package directly, inside Seurat it uses it, and need to increase the size for future. For example, I got the error: Computing within dataset neighborhoods Error in getGlobalsAndPackages(expr, envir = envir, globals = globals) : The total size of the 7 globals exported for future expression (‘FUN()’) is 670.56 MiB.. This exceeds the maximum allowed size of 500.00 MiB (option ‘future.globals.maxSize’). The three largest globals are ‘query’ (670.45 MiB of class ‘numeric’), ‘index’ (85.51 KiB of class ‘S4’) and ‘FUN’ (29.86 KiB of class ‘function’) Calls: IntegrateLayers … getGlobalsAndPackagesXApply -> getGlobalsAndPackages In addition: Warning messages: 1: In asMethod(object) : sparse->dense coercion: allocating vector of size 5.2 GiB 2: In asMethod(object) : sparse->dense coercion: allocating vector of size 2.5 GiB

Other packages I use for integration

  • R: install.packages(c(“here”, “logger”, “import”, “withr”))

Installation of Seurat v5 in mac

I have troubles when install BPCells using conda under MacOS system. So I use brew system to install Seurat in mac. See ref.

  1. Clean the conda effect: quite even base env.
    • It may introduce dependencies errors later.
  2. brew install hdf5 llvm
  3. Then I can install BPCells in R: remotes::install_github("bnprks/BPCells")
  4. R: install.packages(c("R.utils", "future", "remotes"))
  5. R: install.packages("RCurl")
    • have to conda deactivate, and quite base env. otherwise, will have errors.
    • this might because I use brew system to install R
  6. R: install.packages("RSamtools")
  7. R: remotes::install_github("satijalab/seurat-data", "seurat5") remotes::install_github("stuart-lab/signac", "seurat5") remotes::install_github("satijalab/seurat", "seurat5")
  8. R: remotes::install_github('satijalab/seurat-wrappers', ref = "seurat5", force = TRUE, upgrade = "never") BiocManager::install("batchelor") install.packages("harmony")

Working directory in TSCC

  • /projects/ps-renlab2/szu/projects/CEMBA2

Seurat data

  • Under directory out
    • transferLabel: Seurat generated under encoder machine.
    • [deprecated] transferLabel_encoder: a copy of transferLabel
    • transferLabel_tscc: Seurat generated under tscc