December 11, 2018, Tuesday

• For workflow jobs API
  • Add name and description so workflow can be submitted, executed, and saved
  • Don't need to share job with the maizecode user since we will use the maizecode user to call API
  • Workflows are saved into the de.sciapps.org database, then dumped into the public site
• Steps for testing with the maizecode RNAseq data
  • Build workflow on sciapps.org (QC+quant) for two reps
  • Get a token from de.sciapps.org and run on one tissue (two reps)
  • Submit the workflow with the maizecode user by providing the _derived_from with new data
    • Submit, run, and save
    • Need to provide name for the workflow, such as MC_B73_Root_RNAseq
    • Add workflow id to the metadata of four input files in the Data Store
      • Combine metadata from the folder to the input files
      • Do we need a separate token to access Data Store?
      • How to add the SRA accession number into the metadata?
  • Dump the workflow json and save into the public site
    • For de.sciapps.org, call API to get the json file
    • For www.sciapps.org, set up a remote user to save the workflow json into the database
      • Need to update the public site to list new workflows at the beginning, or add time for sorting?
  • Release the Data Store folder
    • Make a 'released' folder under the maizecode folder
    • Make a 'to-be-released' folder under the maizecode folder
      • Move the data to the 'released' folder
      • Run QC+quant pipeline
• EMS workflow
  • Renamed chromosome name as 1,2,.. from chr1, chr2
    • Yinping didn't do it
    • We did it for being consistent with the annotation file (for viewing in the Genome Browser)
  • bcftools view vs call (new version?)
    • Too slow, think about parallelizing it with launcher
  • Keep contigs besides the 10 chromosomes?
  • Test reproducibility with the small datasets