-
Python 3.9.12.
-
Python packages listed in requirements.
Use conda env create -f magpie.yml
to create the conda environment is recommanded.
-
SpliceAI (Use
conda env create -f spliceai.yml
to create the conda environment is recommanded). -
MATLAB CLI.
-
AnnoVar (register required).
-
OMIM database(application required).
Reminder: running MAGPIE on single CPU may take some time because single process of autoFE.
MAGPIE supports variants in CSV format as input. The input file should contain at least 5 columns in the header as follows. Sample file
Chr | Start | End | Ref | Alt | ... |
---|
- Install packages listed in requirements or use magpie conda environment.
- Run
source magpie.sh --mode pred --test_file [filepath] --file_state annotated --visualization
e.g.source magpie.sh --mode pred --test_file data/datasets/test.csv --file_state annotated --visualization
- Run
source magpie.sh --mode pred --test_file [filepath] --file_state unannotated --visualization
e.g.source magpie.sh --mode pred --test_file data/datasets/test.csv --file_state unannotated --visualization
Results would be saved in data/result
.
-
Download and decompress required database for annotating using
bash download.sh
. -
Apply for AnnoVar access, and place all execuatable annotation tools in
./annovar
. -
Apply for OMIM database access, and place 'genemap2.txt' in
data/annotation_database
. -
Install packages and pieces of software manually or use
Dockerfile
to create and run a docker image.docker build -t magpie . docker run -it magpie
-
Run
source magpie.sh --mode train --input_file [filepath]
e.g.source magpie.sh --mode train --input_file data/datasets/denovo.csv
. Model would be saved inoutput/result/
.
Please feel free to contact me(yichengliu at zju.edu.cn) for technical support.