workflow_localfiles.yml

# These files contain URLs to download input files
# curl reads them with -K option. 
# Those can contain parameters and protocols which curl supports.
# You can change path to a local file, which points to the urls found 
# in the comments below.
fastq_files:
  - class: File
    path: /mnt/input/U5c_CCGTCC_L001_R1_001.fastq.gz
  - class: File
    path: /mnt/input/U5c_CCGTCC_L001_R2_001.fastq.gz
# ftp://ftp-trace.ncbi.nih.gov/giab/ftp/data/NA12878/NIST_NA12878_HG001_HiSeq_300x/140407_D00360_0017_BH947YADXX/Project_RM8398/Sample_U5c/U5c_CCGTCC_L001_R1_001.fastq.gz
# ftp://ftp-trace.ncbi.nih.gov/giab/ftp/data/NA12878/NIST_NA12878_HG001_HiSeq_300x/140407_D00360_0017_BH947YADXX/Project_RM8398/Sample_U5c/U5c_CCGTCC_L001_R2_001.fastq.gz 

reference_genome:
  - class: File
    path: /mnt/input/hs37d5.fa.gz
# ftp://ftp.1000genomes.ebi.ac.uk/vol1/ftp/technical/reference/phase2_reference_assembly_sequence/hs37d5.fa.gz

known_indels_file:
  class: File
  path: /mnt/input/Mills_and_1000G_gold_standard.indels.b37.vcf
# ftp://ftp.broadinstitute.org/bundle/b37/Mills_and_1000G_gold_standard.indels.b37.vcf.gz

known_sites_file:
  class: File
  path: /mnt/input/dbsnp_138.b37.vcf.gz
# ftp://ftp.broadinstitute.org/bundle/b37/dbsnp_138.b37.vcf.gz


# this is for BWA. This string was used at BSC for the same pipeline
readgroup_str: '@RG\tID:Seq01p\tSM:Seq01\tPL:ILLUMINA\tPI:330'
# GATK Haplotype caller sequence
sample_name: 'abc2'
chromosome: '1'