- A pipeline is used to postprocess S-PrediXcan/S-TissueXcan pipeline's results
git clone https://github.com/jiamaozheng/S-PrediXcan_Postprocessing_Pipeline- Prepare the following tools or files:
- Outputs from S-PrediXcan or S-TissueXcan pipeline (required)
- Prediction models (optional, only for analyzing S-PrediXcan output)
gwas_snp.txtfile (tab-delimited format) prepared according to instructions and a concrete example (optional, only for generating locuszoom plots)- plink (optional, only for generating locuszoom plots)
- Download LocusZoom (37.5GB)(optional, only for generating locuszoom plots)
| Argument | Abbre | Required | Default | Description |
|---|---|---|---|---|
| --project_name | -p | Yes | None | project name (e.g ovarian_cancer) |
| --metaxcan_folder | -f | Yes | None | file path to metaxcan outputs |
| --models_folder | -d | Yes | '../data/models/' | file path to prediction models |
| --tools_folder | -t | No | '../data/tools/' | plink or gwas snps file |
| --multi_tissue | -m | No | 'false' | "true" for multi_tissue |
| --locuszoom | -l | No | 'true' | "false" for not locuszoom |
Example 1: To analyze outputs from S-PrediXcan pipeline
python run.py -p breast_cancer -f metaxcan_results_folder -d models_folder Example 2: To analyze outputs from S-PrediXcan pipeline without locuzoom plot analysis
python run.py -p breast_cancer -f metaxcan_results_folder -d models_folder -l false Example 3: To analyze outputs from Multi_tissue S-PrediXcan pipeline
python run.py -p breast_cancer -f metaxcan_results_folder -d models_folder -m true Example 4: To analyze outputs from Multi_tissue S-PrediXcan pipeline without locuzoom plot analysis
python run.py -p breast_cancer -f metaxcan_results_folder -d models_folder -m true -l false ../log/- logs../out/- tables and figures- annotation
.csv - top sigificant genes
.csv - top sigificant genes with snps
.csv - manhattan_plot
.pdf - qq_plot
.pdf - region_plot
.pdf - bubble_plot
.pdf - locuszoom_plot
.pdf
- annotation