snakemake

Best Practices for MSI evaluation using MSIsensor-pro

We recommend using Snakemake to run the MSIsensor-pro pipeline, especially when you have a large sample size. For both tumor-only and tumor-normal modes, please first modify the parameters for running MSIsensor-pro in the config.yaml and *.smk files. Then, run your pipeline using Snakemake.

1. configure the config.yaml file

• config.yaml of tumor-normal paired model.

  msisensor-pro: # path of the msisensor-pro
    /path/of/msisensor-pro/cpp/msisensor-pro
  dir_work: # path of your result output
    /path/of/your/pipeline_output/tumor_normal/
  reference: # path of your reference genome
    /path/of/reference/reference.fa
  genome_version: # version ID 
    GRCh38
  sample_path: #path of the sample path, see the  template at demo/data_input/data4test/tumor_normal_paired.samples.info.tsv
    /path/of/tumor_normal_paired/tumor_normal_paired.samples.info.tsv
  

• config.yaml of tumor-only paired model.

   msisensor-pro: # path of the msisensor-pro
     /path/of/msisensor-pro/cpp/msisensor-pro
   dir_work:
    /path/of/your/pipeline_output/tumor_only/
   reference: # path of your reference genome
    /path/of/reference/reference.fa
   genome_version:
     GRCh38
   samples4baseline: #path of the sample info for baseline building, see template at demo/data_input/data4baseline/sample4baseline_info.tsv
     /path/of/tumor_only/data4baseline/sample4baseline_info.tsv
   samples4test: #path of the sample info for MSI evaluating, see template at demo/data_input/data4test/tumor_only.samples.info.tsv
     /path/of/tumor_only/data4test/tumor_only.samples.info.ts