A package to process RNA-seq split-read counts for splicing and generate tissue-specific splicing annotations (SpliceMaps) that quantify tissue-specific splice site usage and isoform competition. SpliceMaps can be generated from RNA-seq split read counts (e.g. as provided by the software FRASER). The required input format is a table of split read-counts, with columns: 'Chromosome', 'Start', 'End', 'Strand', 'samples':
Chromosome | Start | End | Strand | individual 1 | individual 2 | individual 3 |
---|---|---|---|---|---|---|
1 | 17729 | 17733 | + | 9 | 0 | 0 |
1 | 30667 | 30975 | + | 8 | 1 | 7 |
1 | 135802 | 137620 | + | 0 | 2 | 2 |
1 | 320653 | 320880 | + | 1 | 1 | 4 |
1 | 320653 | 324287 | + | 0 | 2 | 8 |
1 | 320938 | 321031 | + | 2 | 4 | 5 |
1 | 320938 | 322037 | + | 8 | 5 | 4 |
1 | 322228 | 324287 | + | 53 | 27 | 40 |
1 | 324345 | 324438 | + | 99 | 54 | 101 |
1 | 324686 | 324718 | + | 0 | 3 | 8 |
SpliceMaps contain the following information:
column name | Description |
---|---|
junctions | Intron with: chrom:donor site:acceptor site:strand |
gene_id | Ensembl gene id |
gene_name | Gene symbol |
gene_type | Gene type |
gene_tpm | Median expression of gene |
transcript_id | List of transcripts that intron is part of |
Chromosome | Chromosome |
Start | Donor site of intron |
End | Acceptor site of intron |
Strand | Strand of intron |
events | List of competing introns with shared splice site |
splice_site | Fixed splice site in splicing event |
ref_psi | Reference level of PSI |
k | Sum of split-read counts supporting the intron (based on cohort that SpliceMap was computed on) |
n | Sum of split-read counts supporting the splice-site (based on cohort that SpliceMap was computed on) |
median_n | Median of n across individuals |
novel_junction | Intron is not annotated in Gencode (binary) |
weak_site_donor | Donor site is not annotated in Gencode (binary) |
weak_site_acceptor | Acceptor site is not annotated in Gencode (binary) |
Clone git repository of splicemap:
git clone git@github.com:gagneurlab/splicemap.git
cd into repo directory:
cd splicemap
Install conda environment:
# Recommended if you have mamba installed
mamba env create -f environment.yaml
# otherwise
conda env create -f environment.yaml
Activate conda environment:
conda activate splicemap
Check notebooks/example.ipynb
Precomputed SpliceMaps can be downloaded from Zenodo for hg19 and hg38.
splicemap_download --version {version} --splicemap_dir {output_dir}
Supported versions: 'gtex_v8' (hg38) and 'gtex_v7' (hg19).
The source code to create and use SpliceMaps is under MIT license. Pre-computed SpliceMaps are under MIT license.