GCViT

About

GCViT is a tool for whole genome visualization of resequencing or SNP array data, which reads data in GFF and VCF format and allows a user to compare two or more accessions to visually identify regions of similarity and difference across the reference genome. Access to data sets can be controlled through authentication.

GCViT is built on top of CViTjs, a Javascript application for viewing genomic features at the whole-genome scale. GCViT is implemented in Go. A Docker image is available. GCViT exposes an API, and can be installed as a server only, with no UI.

Figure 1. An example of haplotype comparisons of 6 soybean accessions.

Explore Soybean SNP data in GCViT (at SoyBase)

Getting Started

An example soybean dataset has been provided to test cvitjs. To get started, it is recommended that you use Docker.

docker run -d -p 8080:8080 legumefederation/gcvit:v1.0.0

GCViT should now be available at http://localhost:8080.

Setup

Setting up a local GCViT instance with your own data requires:

Preparing the data
Configuring the UI
Running GCViT

Preparing the data

Reference Genome Assembly Backbone

Add a GFF3 file that defines the chromosomes for the genome assembly backbone to the ui/cvitjs/data/ folder.

An example GFF3 file is included at ui/cvitjs/data/soySnp/gm_backbone.gff.
Edit ui/cvitjs/cvit.conf, linking the GFF specified in #1 to CViTjs, and defining which CViTjs UI configuration file to use (described in CViTjs documentation).

Genotype Data Sets

Add one or more (optionally gzipped) VCF files representing genotype data to the assets/ directory.

An example dataset is included at assets/SoySNP50k_TestFile_named.vcf.gz

Configuring the API Service

Add a stanza to config/assetsconfig.yaml for each genotype data set (VCF) file to serve it from the API server, and configure other options described below.

Example config/assetsconfig.yaml

server:
  port: 8080
  portTLS: 8888
  certFile: config/testcert.cert
  keyFile: config/testcert.key
  apiOnly: False
  source: gcvit
  binSize: 500000

users:
  username : password

snptestLegacy:
  location: assets/SoySNP50k_TestFile_named.vcf.gz 
  name: soySNP 50k subset [named]
  format: vcf
  restricted:
    - username

The format for each VCF stanza (snptestLegacy in the above example) is as follows:

key: #user-defined unique name for internal key for API requests
  location: relative to root of server directory
  name: display name for dropdowns
  format: vcf (only option for now, automatically checks if gzipped)
  restricted: [optional] whitelist of users that may access this dataset, if not present, data may be accessed by anyone
    - username: username that can access this data
    - username2: another user that can access this datta

The server stanza is optional, and supports the following options:

Option	Default	Use
port	8080	Changes the port GCViT listens on for HTTPS traffic. Defaults to 8080 only if no portTLS is provided. Otherwise ignores HTTP traffic.
portTLS	-	Changes the port GCViT listens for HTTPS traffic. No default provided as you need to set your own key/cert.
certFile	-	Cert file for HTTPS. config/testcert.cert is only for testing purposes and not a default.
keyFile	-	Key file for HTTPS. config/testcert.key is only for testing purposes and not a default.
apiOnly	False	If True, only serves the api routes, ignoring the GCViT frontend
source	gcvit	Value for Column 2 of generated gff files from /api/generateGFF
binSize	500000	Default number of bases used for bins

The users stanza is also optional. Use this configuration option to set one-or-more users to password protect datasets. Without proper credentials, users will never be presented with restricted datasets when using the gcvit ui. The format is one-or more <username> : <password> pairs. Note this only uses BasicAuth headers, and isn't intended to be very secure. Future updates may include better practices if demand is present.

Configuring the UI (optional)

Glyphs for Haplotype Block, Heatmap and Histogram can be customized in ui/gcvit/src/Components/[HaploConfig.js|HeatConfig.js|HistConfig.js] respectively.
Popover The box that pops up when clicking on a glyph in the image can be customized by editing ui/cvit/src/templates/Popover.js
Help box The text for various in-app help can be customized by editing the appropriate file in ui/gcvit/src/Components/HelpTopcs/
CViTjs display options Title, colors, fonts, bin size, ruler tic interval, and other CViTjs display options are defined in CViTjs configuration file (the default is ui/cvitjs/data/soySnp/test-42219.conf, as spedified in ui/cvitjs/cvit.conf). For more information on configuring the CViTjs component of GCViT, please see the documentation here and the example file
- Configuration settings in ui/gcvit/src/Components/DefaultConfiguration.js override CViTjs equivalent configuration settings

Running GCViT using Docker

The easiest way run a local GCViT instance is by using Docker.

Setting the environment variables DOCKER_BUILDKIT=1 to enable BuildKit is recommended for faster, more efficient builds.

To build CViTjs with any UI customizations, the GCViT API server, and start GCViT, execute the following command in the same directory as docker-compose.yml (i.e., in the root of the gcvit git working tree):

docker compose up --build -d

If you wish to build the api with BasicAuth, append --build-arg apiauth=true to the above build command.

The GCViT UI is then accessible via web browser http://localhost:3000, while the GCViT API server is accessible at http://localhost:8080.

Any changes to the files on the host in assets/, data/, ui/cvitjs/data, gcvit/src, or the ui/cvitjs/cvit.conf file will be immediately reflected in the browser. Changes to any other files will require rebuilding the container images and restarting the containers (docker compose up --build -d).

To stop the GCViT service:

docker compose down

Running GCViT on a Production Server

Docker

To deploy a complete container image (UI + API, including the contents of assets/ and config/) of GCViT in production:

(optional) Set the Docker Context to the production host (default localhost).
Build the complete container image on the host specified by the Docker context:

docker compose -f docker-compose.prod.yml build

Deploy:

docker compose -f docker-compose.prod.yml up -d

The GCViT UI (and API) will then be available at http://:8080 , where hostname is the host running the Docker Engine specified by the Docker context.

Alternatively, a complete container image can be built directly with docker build:

docker build -t gcvit . -f Dockerfile

This command will produce a image with the tag of gcvit:1.0 that can be used to build a container. If you want to save time with automated builds and only need the server API component, the build-arg:

--build-arg apionly=false

is provided to skip over the building of the UI components. Similarly, if you wish to build the tool with BasicAuth the build-arg:

--build-arg apiauth=true

Then to deploy using docker run:

docker run -d -p 8080:8080 gcvit

Go + Node Setup

GCViT may also be built and served directly using Go and and Node together. Before beginning, check that Go is set up and Node is configured to at least the most current LTS version (currently 12.14.0).

Building the backend component with Go

The following packages are used with this service:

github.com/awilkey/bio-format-tools-go/gff 
github.com/awilkey/bio-format-tools-go/vcf
github.com/buaazp/fasthttprouter v0.1.1
github.com/fsnotify/fsnotify v1.4.7
github.com/spf13/pflag v1.0.5
github.com/spf13/viper v1.6.2
github.com/valyala/fasthttp

There are several ways to obtain the dependencys, but the reccomended way is to run go get before running the software for the first time. After dependencies are satisfied you can use the internal go development server to test using:

go run server.go

from the /api directory of the project.

By default this will listen on port 8080, but you can change this in the configuration file.

Once up, you can test that the server is up using wget or curl:

protocol	request
wget	wget localhost:8080/api/experiment
curl	curl localhost:8080/api/experiment

Either protocol should return a JSON object that contains a list of all the configured VCF files.

If you wish to compile the server, Go has a robust set of tools for building and cross-compiling binaries. In the most simple form:

go build -o server .

Will builds a binary that has statically linked libraries, making it portable.

If running on a server without a Go compiler configured, the language has support built-in for cross compiling built in. See HERE for details.

Building the frontend components with Node

This repository contains a pre-built version of both the UI component and CViTjs, so this section is mostly optional.

If you wish to add configuration data for CViT without re-building the tool, you may place the files directly in ui/gcvit/build/cvitjs.

If you want to use a custom build of CViTjs, navigate to /ui/cvitjs/ and write and build your changes. This is most often used when wanting to change CViT's css or the click on feature Popover display. Place the resulting files from the build directory directly into ui/gcvit/build/cvitjs/build to test. These changes will last until you rebuild the GCViT ui component.

To build the UI component of CViT:

#first time only
npm run install

#normal build
npm run build

To save changes to CViT, place the build compontent in ui/gcvit/public/cvitjs and rebuild the gcvit component.

To rebuild the UI component of GCViT:

#first time only
npm install

#normal build
npm run build

#with basic authentication elements
npm run buildauth

This will create a webpacked version of the GCViT UI. Most common reasons to rebuild is updating the Help documentation and updating the CSS.

API:

The following API is served by the GCViT service component:

Path	Verb	Returns
/api/experiment	GET	JSON representation of all experiments in assetconfig.yaml
/api/experiment/{experiment}	GET	JSON representation of all PIs in VCF header
/api/generateGFF	POST	returns gff. Expected parameters of Ref={experiment:PI}&Variant={sameexperiment:PI}, with any number of variants

/	GET	tool UI - Only if apiOnly is False
/login	GET	Attempts to authenticate a username and password. Returns status 200 if OK, 401 if not.

Authentication:

To control access to data sets, create users and restrict access, see assestconfig.yaml. The files config/testcert.* are an example of a self-signed SSL certificate. Instruction for generating a new one and be found here

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

README.md

README.md

GCViT

Table of Contents

About

Getting Started

Setup

Preparing the data

Reference Genome Assembly Backbone

Genotype Data Sets

Configuring the API Service

Configuring the UI (optional)

Running GCViT using Docker

Running GCViT on a Production Server

Docker

Go + Node Setup

Building the backend component with Go

Building the frontend components with Node

API:

Authentication:

Files

README.md

Latest commit

History

README.md

File metadata and controls

GCViT

Table of Contents

About

Getting Started

Setup

Preparing the data

Reference Genome Assembly Backbone

Genotype Data Sets

Configuring the API Service

Configuring the UI (optional)

Running GCViT using Docker

Running GCViT on a Production Server

Docker

Go + Node Setup

Building the backend component with Go

Building the frontend components with Node

API:

Authentication: