minerva-cli 0.0.3

Minerva Command Line Interface

Minerva CLI

Introduction

Minerva Command Line Interface allows uploading and importing microscopy images into Minerva Cloud. Images can also be exported out of Minerva Cloud and saved to local disk in OME-TIFF format.

Configuration

Minerva will by default look for a file named .minerva in the user's home directory and load configuration from that file. Minerva needs the configuration to be able to connect to the correct Minerva service.

Arguments given in command line will override values set in the config file. Config file location may be overridden with argument --config [PATH]

Minerva CLI can be setup by running the command minerva configure Minerva CLI will then prompt you for the following pieces of information:

Parameter name	Description
MINERVA_REGION	us-east-1 (aws region where Minerva is installed)
MINERVA_USERNAME	Minerva Username (can be left empty)
MINERVA_PASSWORD	Minerva Password (can be left empty)
MINERVA_ENDPOINT	API Gateway stage URL, e.g. https://xxxxxxxxxx.execute-api.us-east-1.amazonaws.com/dev
MINERVA_CLIENT_ID	Cognito App client id for the user pool

There is also an example config file provided in case it's easier to edit the file. Copy .minerva.example as $HOME/.minerva, and open the file with a text editor to edit values. All the parameter values can also be set with environment variables.

Show help

python minerva.py

Import a directory of images into Minerva Cloud

Replace REPOSITORY_NAME with a repository name, and PATH_TO_DIRECTORY with a path to a directory. All images from the directory and its subdirectories will be imported.

python minerva.py import -r REPOSITORY_NAME -d PATH_TO_DIRECTORY

Import a single file.

python minerva.py import -r REPOSITORY_NAME -f PATH_TO_FILE```

When importing OME-TIFFs, the parameter --local will process the image locally, and in general will make the import faster with only one or few images.

python minerva.py import -r REPOSITORY_NAME -f PATH_TO_FILE --local```

Export OME-TIFF from Minerva Cloud to local disk

The following command will export and save the image by its default name, and save only the highest pyramid level.

python minerva.py export --id IMAGE_UUID

Running on O2

Installation

# Log in O2 with SSH
# Clone the minerva-cli repository
git clone https://github.com/labsyspharm/minerva-cli.git

# Load necessary modules (you may need to restart bash shell after running this the first time)
module load conda2

# Create conda environment
conda create --name minerva python=3.6
conda activate minerva
pip install -r requirements.txt

# Configure minerva.config
cp .minerva.example ~/.minerva
# Edit the values in config file
nano minerva.config
# It's a good idea to prevent other users from reading your configuration file
chmod 700 ~/.minerva

Import images from ImStor into Minerva Cloud

# Log in O2 transfer node with SSH (replace ecommonsid with your eCommons user id)
ssh [e-mail]
# Copy images from ImStor to scratch space, e.g.
mkdir -p /n/scratch3/$USER/dataset
cp /n/files/ImStor/.../image.ome.tif /n/scratch3/$USER/dataset

# Log in O2 with SSH
ssh [e-mail]
# Start interactive O2 session
srun --pty -p interactive --mem 500M -t 0-06:00 /bin/bash

# Activate conda environment
conda activate minerva

# Run Minerva import
# (replace [REPOSITORY] with a repository name)
# All the images from the given directory will be imported
python $HOME/minerva-cli/minerva.py import -r [REPOSITORY] -d /n/scratch3/$USER/dataset