ClinGen Curation Database and Interface

This software creates an object store and user interface for the collection of mappings between human diseases and genetic variation as input by the ClinGen curation staff.

To support multiple versions of node:

First time setup:

    1. Install node version manager (nvm)
    curl -o- https://raw.githubusercontent.com/creationix/nvm/v0.33.0/install.sh | bash
    2. source ~/.bashrc
            Here are the node settings in .bashrc. The install adds it automatically
            export NVM_DIR="$HOME/.nvm"
            [ -s "$NVM_DIR/nvm.sh" ] && \. "$NVM_DIR/nvm.sh"  # This loads nvm
            [ -s "$NVM_DIR/bash_completion" ] && \. "$NVM_DIR/bash_completion"  # This loads nvm bash_completion
    3. nvm --version
            0.35.2
    4. nvm install 10.16
    5. nvm use 10.16

Switchin to 10.16:

    1. source ~/.bashrc
    2. nvm use 10.16

Current Branching and Merge Strategy

Application is divided into to main branches Dev and Master. Dev is the main branch for refactor development and holds the most up-to-date stable code. After Dev branch has been sufficiently tested and deemed ready for production work loads it can be merged with master. Branches for development are created off of dev and named in the following pattern. 'Feature/{CORRESPONDING-JIRA-TICKET}/{DESCRIPTION}'. For example if you were creating a new branch to work on a new feature for variant approvals it might look like this: feature/CLINAWS-73/variant-approval. Ultimately the git strategy used by this project most closely resembles git flow as documented here

Steps to branching and merging

Clone repo
git checkout dev
git checkout -b feature-branch
Do all work here
git commit and git push feature-branch
git merge dev (pull dev into feature-branch for testing)
git push
git checkout dev
git merge feature-branch

Dev Setup

Pre Reqs

Serverless CLI installed on machine

Setting up Backend Dev environment

Clone the GCI-VCI-AWS repository/pull dev Follow the branching strategy above to switch to the feature specific branch for the story you are working on.

Option A: Developing Locally

Follow the README found in gci-vci-serverless to set up serverless offline and dynamodb local. Replace the value in gci-vci-react/config.js with the localhost endpoint started by serverless offline

Option B: Developing against Dev database

From the gci-vci-serverless directory run serverless info --stage dev take the base endpoint URL (everything before /variants in the first line) and place that value in gci-vci-react/config.js

For development Amplify should no longer be needed.

Pipeline Setup

Edit gci-vci-serverless-samplePolicy.json so that it represents the correct stage needed for the pipeline. This will consist of editing the Resources sections to change from gci-vci-* to gci-vci-STAGE
On the AWS Console create a new role that trusts CodeBuild.
Create a new policy to be associated with the role and inpute the json from gci-vci-serverless-samplePolicy.json
On the Console navigate to CodePipeline. Click create new Pipeline
Allow CodePipeline to create a new role for you. This role is used by Code Pipeline and is different from the role we just created which will be used by CodeBuild shortly
Select Source Provider as github and connect to github. It is important that the individual who creates this connection has the proper permissions and will be at ClinGen for the forseeable future, best practice here would be to utilize a organizational account not owned by an individual.
Select the gci-vci-aws repository and select desired branch.
Select change detections as GitHub Webhooks, this will allow AWS to detect changes in your selected branch and automatically build the project when a change is pushed to the branch
Select build tool as AWS CodeBuild and click Create Project
Give your CodeBuild a project name, make sure to include the stage name this build will be for.
Select managed image and an Ubuntu operating system; Runtime Standard; Image Standard4.0; Environment Type Linux
Select existing role and use the role you created in steps 1-3
Select use buildspec file.
Finish up in CodePipeline

pageneck/gene-and-variant-curation-tools