GENIE-Sponsored-Projects

This repository will contain processing code for GENIE sponsored projects used to for creating cBioPortal files

Installation

Using Python

1. Close the repository and navigate to the READMD.md in your local directory

2. Build the environment using pip install -e .

Using Docker

1. Install Docker: https://docs.docker.com/get-docker/

2. Clone this repository and navigate to the directory

git clone [email protected]:Sage-Bionetworks/GENIE-Sponsored-Projects.git
cd GENIE-Sponsored-Projects

3. Build the container locally.

docker build -t geniesp .

Synapse credentials

Cache your Synapse personal access token (PAT) as an environmental variable:

export SYNAPSE_AUTH_TOKEN={your_personal_access_token_here}

Usage

To view usage details, run

docker run --rm geniesp -h

Output will be as follows

usage: geniesp [-h] [--staging]
               {NSCLC,CRC,BrCa,PANC,Prostate,AKT1,ERRB2,FGFR4} release

Run GENIE sponsored projects

positional arguments:
  {NSCLC,CRC,BrCa,PANC,Prostate,AKT1,ERRB2,FGFR4}
                        Specify project to run
  release               Specify bpc release (e.g. 1.1-consortium)

optional arguments:
  -h, --help            Show this help message and exit
  --upload              Upload files into Synapse BPC staging directory. Default: false
  --cbioportal {synapseID}
                        Optional parameter to specify cbioportal folder
                        location
  --production          Whether to run in production mode or not. Default: false

Example command line:

This runs the release pipeline for BLADDER 1.1 in non-production mode (staging).

geniesp BLADDER 1.1-consortium --upload

Example command using docker:

This runs the release pipeline for PANC 1.1 in non-production mode (staging).

docker run --rm -e SYNAPSE_AUTH_TOKEN=$SYNAPSE_AUTH_TOKEN geniesp geniesp PANC 1.1-consortium --upload

Scripts

To validate a cBioPortal mapping file stored on synapse:

python validate_map.py -s {synapse_id} -c {cohort} -r {release} -l info

or stored in a local file:

python validate_map.py -f {/path/to/file.csv} -c {cohort} -r {release} -l info

To view full usage details:

python validate_map.py -h

which outputs:

usage: validate_map.py [-h] [--synapse_id SYNAPSE_ID | --file FILE] [--version VERSION] [--cohort {BLADDER,BRCA,CRC,NSCLC,PANC,PROSTATE}]
                       [--release {1.1-consortium,1.2-consortium,2.0-public,2.1-consortium}] [--outfile OUTFILE] [--log {debug,info,warning,error}]

Checks validity of BPC to cBioPortal mapping file

optional arguments:
  -h, --help            show this help message and exit
  --synapse_id SYNAPSE_ID, -s SYNAPSE_ID
                        Synapse ID of mapping file
  --file FILE, -f FILE  Local path to mapping file
  --version VERSION, -v VERSION
                        Synapse entity version number (default: current)
  --cohort {BLADDER,BRCA,CRC,NSCLC,PANC,PROSTATE}, -c {BLADDER,BRCA,CRC,NSCLC,PANC,PROSTATE}
                        BPC cohort label (default: BLADDER)
  --release {1.1-consortium,1.2-consortium,2.0-public,2.1-consortium}, -r {1.1-consortium,1.2-consortium,2.0-public,2.1-consortium}
                        Release label (default: 1.1-consortium)
  --outfile OUTFILE, -o OUTFILE
                        Name of output file (default: output.csv)
  --log {debug,info,warning,error}, -l {debug,info,warning,error}
                        Set logging output level (default: error)

Troubleshooting

The most common issues when running GENIE-Sponsored-Projects code for BPC involve changes to variable names of the underlying source data and outdated or incorrect references.

Variable references

1. syn22294851: the Scope of Release maintains a running log of changes to derived variable names
2. syn17011602: curated variable names are listed in data dictionaries for each cohort

If obtaining an error that a variable cannot be found, check and update the following references on Synapse:

1. syn25712693: BPC REDCap to cbio mapping
2. syn22296821: Dataset labels in Data files for derived variables

On rare occassions, variable name changes may also require changes in the codebase:

1. Check for any hardcoded variable names have been updated

Finally, some information is not collected for particular cohorts.

1. syn20852283: check if data is collected for a given cohort by investigating raw file uploads