Attention:
The frontline support team will be unavailable to answer questions on April 15th and 17th 2019. We will be back soon after. Thank you for your patience and we apologize for any inconvenience!
Latest Release: 03/12/19
Release Notes can be found here.

Unable to update method config

dheimandheiman Member, Broadie ✭✭
edited September 2017 in Ask the FireCloud Team

When I try to update the optional input aggregate_clusters.mRNAcnmf with the value this.mRNA_protein_coding_CNMF_bestcluster in method config Aggregate_Molecular_Subtype_Clusters in workspace nci-dheiman-b-org/dev, I get the following error:

and I'm returned to the method config's launch screen.

Tagging @mnoble and @smeier so they can track this issue and potentially add users to the method to debug on Friday while I'm out. [email protected] has already been added to this workspace.


In what looks to be a related issue, when I try to launch the method config with what I could configure, I get yet another error:

Best Answer

Answers

  • KateNKateN Cambridge, MAMember, Broadie, Moderator admin

    That does look like a bug to me at first glance. I will ask a developer to take a look and get back to you.

  • mnoblemnoble Broad Institute of MIT & HarvardMember, Broadie

    FWIW: using the attached method config I was able to accomplish what David was trying to do ... and install sans error to nci-dheiman-b-org/mnoble_test workspace.

  • dheimandheiman Member, Broadie ✭✭

    When I used the API to retrieve the config, it looks like somehow the previous version persisted, such that the API returns an amalgam of the old and new versions:

    % fissfc config_get -c Aggregate_Molecular_Subtype_Clusters
    0
    {
      "name": "Aggregate_Molecular_Subtype_Clusters",
      "methodRepoMethod": {
        "methodNamespace": "broadgdac",
        "methodName": "aggregate_clusters",
        "methodVersion": 5
      },
      "outputs": {
        "aggregate_clusters_workflow.aggregate_clusters.aggregate_clusters_pkg": "this.pkg__Aggregate_Molecular_Subtype_Clusters",
        "aggregate_clusters_workflow.aggregate_clusters.aggregate_molecular_clusters": "this.aggregate_molecular_signatures",
        "aggregate_clusters_workflow.aggregate_clusters.aggregate_molecular_clusters_transposed": "this.aggregate_molecular_signatures_trans"
      },
      "inputs": {
        "aggregate_clusters_workflow.aggregate_clusters.miRNAMaturecnmf": "this.miR_CNMF_bestcluster",
        "aggregate_clusters_workflow.aggregate_clusters.mirnacnmf": "",
        "aggregate_clusters_workflow.aggregate_clusters.PredictionU133aTCGA": "",
        "aggregate_clusters_workflow.aggregate_clusters.mergedfile": "this.name",
        "aggregate_clusters_workflow.aggregate_clusters.copynumber": "this.CN_CNMF__bestcluster",
        "aggregate_clusters_workflow.aggregate_clusters.lincRNAcnmf": "this.mRNA_lincRNA_CNMF_bestcluster",
        "aggregate_clusters_workflow.aggregate_clusters.miRseqcnmf": "",
        "aggregate_clusters_workflow.aggregate_clusters.mRNAseqcnmf": "this.mRNA_protein_coding_CNMF_bestcluster",
        "aggregate_clusters_workflow.aggregate_clusters.mrnahc": "",
        "aggregate_clusters_workflow.aggregate_clusters.PredictionU133aPhilips": "",
        "aggregate_clusters_workflow.aggregate_clusters.package": "workspace.package",
        "aggregate_clusters_workflow.aggregate_clusters.mirnahc": "",
        "aggregate_clusters_workflow.aggregate_clusters.methylationhc": "",
        "aggregate_clusters_workflow.aggregate_clusters.PredictionUNCTCGA": "",
        "aggregate_clusters_workflow.aggregate_clusters.mrnacnmf": "",
        "aggregate_clusters_workflow.aggregate_clusters.methylationcnmf": "this.Methylation_CNMF_bestcluster",
        "aggregate_clusters_workflow.aggregate_clusters.mRNAseqhc": "",
        "aggregate_clusters_workflow.aggregate_clusters.miRseqMaturecnmf": "this.miR_CNMF_bestcluster",
        "aggregate_clusters_workflow.aggregate_clusters.miRseqMaturehc": "",
        "aggregate_clusters_workflow.aggregate_clusters.miRseqhc": "",
        "aggregate_clusters_workflow.aggregate_clusters.CustomEvents": "workspace.custom_events_file",
        "aggregate_clusters_workflow.aggregate_clusters.copynumberth": "this.CopyNumber_Clustering_CNMF_thresholded__bestcluster"
      },
      "rootEntityType": "sample_set",
      "prerequisites": {
    
      },
      "methodConfigVersion": 6,
      "deleted": false,
      "namespace": "broadgdac"
    }
    

    After deleting and re-importing the config, it seems to be working.

  • esalinasesalinas BroadMember, Broadie ✭✭✭
    edited October 2017

    @dheiman @KateN I experienced something like this last week. I had a WDL and a task that was un-called in the WDL (just defined). And the system was complaining about an error in that task. SO, to address that error, I edited the WDL and totally removed the task. BUT after updating the WDL with the task totally removed, the system still complained that the task was causing an error. From inspection of the WDL that would be impossible because I did a string search and the name did not appear at all. So that means that somehow, even though the WDL was updated, something somehwere was "stuck" or pointing to a previous WDL.

    what I did to get around the issue was just create a new WDL name to start with snapshot 1 and the error went away. That proves that the latest WDL upload did not have a problem, but something internal was "stuck" to an older WDL.

    Post edited by esalinas on
  • kmegqkmegq BroadMember, Broadie

    I experienced this problem this morning, as well. After updating a WDL to a new snapshot and removing one of the input parameters, I still got an error message that there was an extraneous input. Using the workaround that @esalinas describes (recreating the same method with a new name as snapshot 0) worked.

  • esalinasesalinas BroadMember, Broadie ✭✭✭

    hi @kmegq i now know I'm not the only one to have experienced that issue.

  • dheimandheiman Member, Broadie ✭✭

    @KateN, it sounds like this is a new bug, as I just re-encountered it - removing/renaming parameters in the WDL doesn't quite work anymore. I don't think adding new ones breaks anything, but removed/renamed parameters don't look like they're removed/replaced in the database.

  • KateNKateN Cambridge, MAMember, Broadie, Moderator admin

    Thank you for the updates; A developer is looking at this now.

  • davidandavidan Portland, MEMember, Broadie

    Hi all - I've taken a look at this and verified it is a real bug. It's a new one, as you guessed. We'll start work on fixing it; in the meantime, the best workaround is to create a brand new config pointing at the desired WDL snapshot instead of editing an existing config. I realize that's cumbersome, but hopefully that gets you working until we fix it!

Sign In or Register to comment.