Tutorial

Dependency data download

To predict breast cancer risk with MiScan, Users firstly needs to do:

• VCF Files

  We recommend using whole exon sequencing data to obtain individual variation information, but the results of whole genome sequencing and full-length RNA-seq data are also compatible with the model. Please download test VCF files from FTP

• Maxout model weight

  Users also need a MiScan model weight to perform prediction, well-trained model weight can be downloaded from here

usage-commandline

Suppose your VCF file and weight are placed in the $dir directory, run with code:

MiScan --vcf $dir/SRR5447191.combined.filtered.vcf -o outputs --weight $dir/_MiScan_weights.hdf5

or with below command:

python -m MiScan --vcf $dir/SRR5447191.combined.filtered.vcf -o outputs --weight $dir/_MiScan_weights.hdf5

docker-commandline

if with docker, the path of VCF file or weight path shoule be path in Docker environment:

docker run -v $dir_contains_VCF_MODEL_FILES:/miscan jefferyustc/miscan_command_line:latest \
 -o test_outputs --vcf sample.vcf --weight weights.hdf5

Attention: $dir_contains_VCF_MODEL_FILES should be a dir in your local system that contains sample vcf file and weight file

usage-script

users can also execute this from script:

from MiScan import miscan_main
miscan_main(
    outDir='./outputs',
    inVcf='/Users/jeffery/Downloads/SRR5447191.combined.filtered.vcf',
    model_weight='/Users/jeffery/workspace/projects/outputs/_MiScan_weights.hdf5'
)