December 18, 2018, Tuesday

• MaizeCode metadata example for B73LongRampage
  • PRJNA438108
  • BioProject_PRJNA438108 (sra_project_id: PRJNA438108) --> LongRampage
    • BioSample1 (cultivar: B73; tissue: endosperm; replicate: 1) --> endosperm_rep1
      • BioSample1Library1 (assay: Long RNA-Seq (LG)) --> RNAseq
      • Long RNA-Seq of Maize: B73 Endosperm Rep 1
        • 300282_S1_R1_001.fastq.gz --> endo_rep1_R1.fq.gz
        • 300282_S1_R2_001.fastq.gz --> endo_rep1_R2.fq.gz
      • BioSample1Library2 (assay: RAMPAGE (RM)) --> RAMPAGE
• Goal to combine metadata to one
  • Query SRA database with Sample name and BioProject ID to get SRR number?
    • https://www.biostars.org/p/121762/
    • http://bergmanlab.genetics.uga.edu/?p=1072
    • http://bioinfo.umassmed.edu/bootstrappers/guides/main/python_get_sra_run_ids.html
  • Add SRR number to the metadata so that we can pull the data from NCBI
• Modifications to the public site
  • Change 'Load a workflow' page to handle loading multiple workflows
  • So that we can load multiple workflows and perform 'combinational' downstream analysis
• Steps for running through API
  • tok="Authorization: Bearer 77f32c7173ebe6578df669ca449b0"
  • curl -X POST -sk -H "user: lwang" -H "$tok" https://de.sciapps.org/workflowJob/new?runWorkflowJob=1 -F "fileToUpload=@test2.json" > testw.json
    • {
    • "wf_step_0:trait": "agave://data.iplantcollaborative.org/lwang/sci_data/height.txt",
    • "_derived_from": "5fb15c8f-0d38-4ec3-9b0e-392b59f08883"
    • }
  • To get workflow_id returned
    • jq '.data.workflow_id' testw.json
  • To get cleaned workflow json
    • jq -c '.data.workflow' testw.json
  • Create my.sql file
    • insert into workflow (workflow_id, name, description, json) values("0a958127-55fc-48ea-8574-9e6b8185a65d", "testmaizecode", "test", 'workflow_json_output_from_above')
  • mysql -u sciapps -p$PWD sciapps < my.sql (on halcott)
    • Jobs are missing on the public site unless the job json file is grabbed after jobs are submitted to Agave, which won't take more than 5 minutes since MCrna app is simplified, ideally we will add API to the public site
      • curl -X GET -sk -H "user: lwang" -H "$tok" https://de.sciapps.org/workflow/0a958127-55fc-48ea-8574-9e6b8185a65d > testw.json
      • jq -c '.data' testw.json
      • Create my.sql file
        • insert into workflow (workflow_id, name, description, json) values("0a958127-55fc-48ea-8574-9e6b8185a65d", "testmaizecode", "test", 'workflow_json_output_from_above')
      • mysql -u sciapps -p$PWD sciapps -e'delete from workflow where workflow_id="0a958127-55fc-48ea-8574-9e6b8185a65d"'
      • mysql -u sciapps -p$PWD sciapps < my.sql (on halcott)
      • https://www.sciapps.org/?wf_id=0a958127-55fc-48ea-8574-9e6b8185a65d
        • Need to share the job with maizecode, unless the job is submitted by maizecode