Process and integrate 10x Multiome Data
The following scripts should be used in this order to process the multiome data:
- Create conda environment with this yml file: environment_signac.yml
- Install packages in R (within the conda environment) as needed: 99_install_signac_packages.R
Script: 01_multiome_signac.R
Input arguments:
- argument 1: cellranger out path
- argument 2: sample name
- argument 3: species, human or mouse
Example Usage:
Rscript ~/scripts/01_multiome_signac.R /athena/melnicklab/users/scratch/chc2077/ls_multiome/ls_multiome/outs ls_multiome mouse
Object will be saved into the cellranger-arc output folder
Create a csv runsheet that has a unique name for each sample, and a path to cellranger-arc "out" folder
csv runsheet header must contain "Sample" "sobj_path". Additional columns will be added as metadata
Example runsheet included: Example_runsheet.csv
Script: 21_multiome_velo.R
This script will add the velocity data to multiome objects. Argument is a path to the runsheet
This script will create multiome objects if 01_multiome_signac.R was not run, but will not contain velocity data
Script: 08_combined_multiome_2.R
Input arguments:
- argument 1: path to runsheet
- argument 2: species to use (Human or Mouse)
- argument 3 (optional): number of cores to use
Example Usage:
Rscript ~/.scripts/08_combine_multiome.R ./ls_msobj_combine_runsheet.csv mouse 32
A work directory will be created within the directory that this script is run from. The combined object will be saved as "combined.multiome.sobj.rds" within this work folder
Script: 12_multiome_harmony_rds.R
Input arguments:
- argument 1: path to object - use the "combined.multiome.sobj.rds" file generated by 08_combined_multiome_2.R
- argument 2: object name - name to give the object
- argument 3: batch name - name of the metadata column to correct for. orig.ident
- argument 4 (optional): number of cores to use
Example Usage:
Rscript ~/scripts/12_multiome_harmony_rds.R ./work/integrated.multiome.sobj.celltype.fine.celltype.rds ls_am_dm_multiome orig.ident
Script: 16_label_singlecell_3.R
- argument 1: Path to seurat or signac object as an RDS file
- argument 2: Path to reference seurat or signac object as an RDS file
- argument 3: Name of assay to use
- argument 4+: Names of labeles to transfer
Example Usage:
Rscript ~/scripts/16_label_singlecell.R ./work/integrated.multiome.sobj.rds /athena/masonlab/scratch/users/chc2077/references/gcb/20210512_melnick.int.sobj.RDS RNA celltype fine.celltype