mandana-mazaheri / conp-recommender Goto Github PK

CONP contains a diverse range of 61 tools and pipelines (June 2020). Many of these tools/pipelines come from neuroscience or genomics research institutes.
Requirements:
CONP pipelines can be easily installed and run on any computer with a container engine. You will need to install:

• Docker or Singularity
• Boutiques python package (pip install boutiques)

How you can run a CONP tool/pipeline?
First you need to ensure you have boutiques installed with Docker or Singularity.

Tools/pipelines can be run on Boutiques based on their Zenodo ID as follows:

1. The example command will create a first minimal invocation so that you don't have to start from scratch: bosh example <zenodo_id>. If you feel like starting with a more complete set of options, you can pass --complete to the example command: bosh example --complete <zenodo_id>

2. You can now edit a JSON example invocation (example_invocation.json) to add your input values based on the example provided in 1.

3. You are now all set to use the exec command to launch an analysis. Note, the option -s will show live stdout/stderr in the terminal.

    ` bosh exec launch -s <zenodo_id> <invocation>`
   

Instruction to bosh publish:
The details of each execution are captured and recorded in a publicly safe format so that you can review past analysis runs. These records are stored in the Boutiques cache and capture each executions' descriptor, invocation and output results. Input and output file hashes are included to easily compare results between different analyses. bosh data is the command to interact and publish execution records. For instance, here is the record for one of your runs:

%%bash
bosh data inspect -e

Don't worry though, nothing gets public until you explicitly do so by running bosh data publish:

%%bash
bosh data publish --sandbox -y

A publisher of Boutiques tools in Zenodo (http://zenodo.org (http://zenodo.org/)). Requires a Zenodo access token, see http://developers.zenodo.org/#authentication (http://developers.zenodo.org/#authentication).

usage: bosh publish [-h] [--sandbox] [--zenodo-token ZENODO_TOKEN] [--no-int]
                    [-v] [-r | --id ID]
                    boutiques_descriptor

• For install a data-set from CONP portal, you can follow this issue.

• In case of any error, you can see CONP tutorials for additional information.

• If you are new to boutiques, we recommend following boutiques tutorial.

To download a specific CONP dataset you should follow these steps:
1- Install Datalad.
2- Install the git repository of CONP by datalad using this command:

datalad install https://github.com/CONP-PCNO/conp-dataset.git

3- To install the dataset, go into the created conp-dataset directory and run:

cd conp-dataset
datalad install projects/THE-NAME-OF-DATASET-YOU-WANT

Now that the DataLad dataset has been installed, go into the dataset directory under
projects/THE-NAME-OF-DATASET-YOU-WANT
cd projects/THE-NAME-OF-DATASET-YOU-WANT

Note that files visible in the dataset are symlinks and will need to be downloaded manually using the datalad get command in the dataset directory:

datalad get <filepath>

Note, if you run datalad get * command, all the files present in the dataset directory will be downloaded.