Differential RNA splicing analysis with single-cell/nucleus RNA-seq data

Note

You need to install a Docker image of Shiba (and clone the Shiba GitHub repository to run SnakeScShiba). If you don't have them installed, please follow the instructions in the Installation section.

---

Before you start

• Perform mapping of sc(sn)RNA-seq reads to the reference genome using STARsolo.
• Download a gene annotataion file of your interest in GTF format.

Here is an example code for downloading a mouse gene annotation file (Ensembl 102):

wget https://ftp.ensembl.org/pub/release-102/gtf/mus_musculus/Mus_musculus.GRCm38.102.gtf.gz
gzip -d Mus_musculus.GRCm38.102.gtf.gz

---

scShiba

1. Prepare inputs

experiment.tsv: A tab-separated text file of barcode file and STAR solo raw output directory.

barcode<tab>SJ
/path/to/barcodes_run1.tsv<tab>/path/to/run1/Solo.out/SJ/raw
/path/to/barcodes_run2.tsv<tab>/path/to/run2/Solo.out/SJ/raw
/path/to/barcodes_run3.tsv<tab>/path/to/run3/Solo.out/SJ/raw
/path/to/barcodes_run4.tsv<tab>/path/to/run4/Solo.out/SJ/raw

barcodes.tsv is a tab-separated text file of barcode and group name like this:

barcode<tab>group
TTTGTTGTCCACACCT<tab>Cluster-1
TCAAGACCACTACAGT<tab>Cluster-1
TATTTCGGTACAGTAA<tab>Cluster-1
ATCCTATGTTAATCGC<tab>Cluster-1
ATCGATGAGTTTCTTC<tab>Cluster-2
ATCGATGGTCTTGCTC<tab>Cluster-2
TATGTTCGTCAGGCAA<tab>Cluster-2
ATCGCCTAGACTCGAG<tab>Cluster-2
...

Please replace <tab> with a tab character.

config.yaml: A yaml file of the configuration.

workdir:
  /path/to/workdir
gtf:
  /path/to/Mus_musculus.GRCm38.102.gtf
experiment_table:
  /path/to/experiment.tsv

# PSI calculation
only_psi:
  False
fdr:
  0.05
delta_psi:
  0.1
reference_group:
  Cluster-1
alternative_group:
  Cluster-2
minimum_reads:
  10
excel:
  True

You can generate a file of splicing analysis results in excel format by setting excel to True.

2. Run

Docker:

cp experiment.tsv config.yaml /path/to/workdir
cd /path/to/workdir
docker run --rm -v $(pwd):$(pwd) naotokubota/shiba:latest \
python /opt_shiba/Shiba/scshiba.py -p 32 /path/to/workdir/config.yaml

Singularity:

cp experiment.tsv config.yaml /path/to/workdir
singularity exec shiba_latest.sif \
python /opt_shiba/Shiba/scshiba.py -p 32 /path/to/workdir/config.yaml

Note

When you use Singularity, you do not need to bind any paths as it automatically binds some paths in the host system to the container. In the default configuration, the system default bind points are $HOME, /sys:/sys, /proc:/proc, /tmp:/tmp, /var/tmp:/var/tmp, /etc/resolv.conf:/etc/resolv.conf, /etc/passwd:/etc/passwd, and $PWD. If files needed to be accessed are not in these paths, you can use the --bind option to bind the files to the container.

---

SnakeScShiba

A snakemake-based workflow of scShiba. This is useful for running scShiba on a cluster. Snakemake automatically parallelizes the jobs and manages the dependencies between them.

1. Prepare inputs

experiment.tsv: A tab-separated text file of sample ID, path to fastq files, and groups for differential analysis. This is the same as the input for scShiba.

config.yaml: A yaml file of the configuration. This is the same as the input for scShiba but with the addition of the container field.

workdir:
  /path/to/workdir
container: # This field is required for SnakeScShiba
  docker://naotokubota/shiba
gtf:
  /path/to/Mus_musculus.GRCm38.102.gtf
experiment_table:
  /path/to/experiment.tsv

# PSI calculation
only_psi:
  False
fdr:
  0.05
delta_psi:
  0.1
reference_group:
  Cluster-1
alternative_group:
  Cluster-2
minimum_reads:
  10
excel:
  False

You can generate a file of splicing analysis results in excel format by setting excel to True.

2. Run

Please make sure that you have installed Snakemake and Singularity and cloned the Shiba repository on your system.

snakemake -s /path/to/Shiba/snakescshiba.smk \
--configfile config.yaml \
--cores 16 \
--use-singularity \
--rerun-incomplete