This project describes an automated pipeline to quantify isoform expressions for the GTEx V8 RNA sequencing data. The protocol uses XAEM, a powerful method for isoform expression estimation across multiple samples. You also can find the detailed information about XAEM website and in the published paper in Bioinformatics.
R (recommended version >= 3.5.1)
Python (recommended version >= 3.7)
conda create -n iGTEx python=3.10 r=4.2 -c conda-forge
conda activate iGTEx
pip install pandas
R -e "install.packages(c('foreach','doParallel'), repos='https://mirrors.tuna.tsinghua.edu.cn/CRAN/')"
Install the iGTEx_XAEM tool for this protocol via the bash command:
wget https://github.com/ZhengCQ/iGTEx/archive/refs/tags/iGTEx_XAEM_v0.1.2.zip
unzip iGTEx_XAEM_v0.1.2.zip
ln -fs iGTEx-iGTEx_XAEM_v0.1.2 iGTEx_XAEM
Run the following commands to download the reference annotating the transcripts:
cd /path/to/iGTEx_XAEM
python down_ref.py
(Optional) To down a particular reference with gencode hg38, use:
python down_ref.py -db gencode_38
An example is prepared in the project Example folder, executable as
cd /path/to/iGTEx_XAEM/Example
sh run_example.sh
XAEM performs better when multiple samples of similar data type are considered. In the GTEx V8 data, for each tissue, we create a project directory:
mkdir -p /path/to/Tissue1
cd /path/to/Tissue1
In /path/to/Tissue1, create a file /path/to/Tissue1/infastq_lst.tsv listing the FASTQ input files. The file is a tab-delimited text file with 4 columns: Sample name, Source name, FASTQ file name for paired-end read 1, and FASTQ file name for paired-end read 2. Source name indicates the batch or sequencing library of the sample, so that the same sample may correspond to more than one sources. A standard example, where each sample has only a single batch or multiple batches, is given as /path/to/iGTEx_XAEM/Example/infastq_lst.tsv:
sample4 S0007 S0007_1.fg.gz S0007_2.fg.gz
sample5 S0008 S0008_1.fg.gz S0008_2.fg.gz
sample1 S0001 S0001_1.fg.gz S0001_2.fg.gz
sample1 S0002 S0002_1.fg.gz S0002_2.fg.gz
sample2 S0003 S0003_1.fg.gz S0003_2.fg.gz
sample2 S0004 S0004_1.fg.gz S0004_2.fg.gz
sample3 S0005 S0005_1.fg.gz S0005_2.fg.gz
sample3 S0006 S0006_1.fg.gz S0006_2.fg.gz
XAEM can be easily run with:
python /path/to/iGTEx_XAEM/run_xaem.py -i /path/to/Tissue1/infastq_lst.tsv
(Optional) To specify a particular reference with gencode hg38, use:
--ref gencode_38
(Optional) To specify a particular output directory, use:
-o /path/to/Tissue1_output_directory
(Optional) Further customized configuration of XAEM can be setup by:
-c /path/to/Tissue1_config.ini
An example of the config.ini file can be found in /path/to/iGTEx_XAEM/.
Isoform ratio can be easily run with
python /path/to/iGTEx_XAEM/exp2ratio.py -i /path/to/XAEM_isoform_expression.RData
(Optional) To specify a particular reference with gencode hg38, use:
--ref gencode_38
(Optional) The default output directory as same as input isoform file. To specify a particular output directory, use:
-o /path/to/Tissue1_output_directory
(Optional) To specify a covariates file for linear regression:
--covariates /path/to/covariates_file
(For GTEx) For GTEx v8, we have prepared the covariates information for calculation, you can specify a tissue name:
--tissue Brain_Amygdala
(For GTEx) For GTEx v8 demo:
python /path/to/iGTEx_XAEM/exp2ratio.py -i /path/to/Brain_Amygdala_XAEM_isoform_expression.RData --ref gencode_38 --tissue Brain_Amygdala
/path/to/osca --efile /path/to/TissueName/isoform_splice_ratio.tsv --gene-expression --make-bod --no-fid --out TissueName
/path/to/osca --befile TissueName --update-opi /path/to/iGTEx_XAEM/ref/gencode_38/anno_gene_info.opi
/path/to/osca --sqtl --bfile /path/to/Genotype/BED_All/TissueName_Genotype --befile TissueName --maf 0.05 --call 0.85 --cis-wind 1000 --thread-num 10 --task-num 1 --task-num 1 --task-id 1 --to-smr --bed /path/to/iGTEx_XAEM/ref/gencode_38/anno_gene_info.bed --out sQTL_results/TissueName