Scout is well integrated with the Chanjo software for coverage reporting. While Chanjo is efficient and speedy in calculating metrics over transcript data, making it a valuable tool for visualizing statistics in WES analyses, the incorporation of support for other types of analyses (WGS, WTS) in Scout has rendered Chanjo outdated. In particular, updating the reference transcript set in an active Chanjo instance is challenging.
In response to this, efforts at Clinical Genomics led to the development of Chanjo2, an updated tool designed to calculate coverage and coverage completeness statistics with the advantage of utilizing the .d4 (Dense Depth Data Dump) format.
For further information on d4tools and the .d4 format, please visit the relevant project pages. Instructions on setting up a Chanjo2 server can be found in the documentation available on the Chanjo2 GitHub pages.
For system administrators configuring communication with either of these tools, we recommend integrating with Chanjo2, as the Chanjo repository is likely to be deprecated.
Integration with Chanjo2 does not require the installation of any specific libraries. However, it needs the availability of a Chanjo2 instance with access to the disk folder where scout case data files are stored.
Once the Chanjo2 server is accessible, simply modify one line in the Scout configuration file to set the value of 'CHANJO2_URL' to the root endpoint of your Chanjo2 API.
To enable a Scout case to display links to Chanjo2 pages such as gene coverage reports and genes coverage overviews, the .yaml case load config file must specify the path to .d4 files for at least one of the case individuals/samples.
Prior to loading a case into Scout, it's essential to create .d4 files. Instructions on how to generate these files using the d4tools software can be found in the software's documentation.
The path to the .d4 files should be provided in the .yaml Scout load config, as outlined in the scout documentation.
Scout may be configured to visualize coverage reports produced by Chanjo. To do so it is necessary to install the python package Chanjo and its report visualization app, chanjo-report.
Both these programs require a database to store samples names, transcript specifications and statistics relative to the trascript coverage for all samples. Note that chanjo can be used as a standalone but chanjo-report needs to have access to the data produced by chanjo to generate coverage reports. Scout does not require chanjo installed in the path to visualize coverage reports, but it is necessary to install chanjo-report and set up the connection to a database containing coverage data in order to use this fuctionality.
Chanjo-report (and chanjo) work with MySQL and SQLite databases. This guide explains how to set up a MySQL database since this database is the one used in production at Clinical Genomics. For a guide on how to install MySQL on your server click here.
Once the database and the administrator user and password are configured, you can create the new database to contain the chanjo structures. To do so, from a terminal type:
mysql -u admin_user -padmin_password #no space between -p and password
mysql> create database chanjo4_demo;
mysql> exitTo create a new empty database download the empty schema chanjo4_structure.sql under scout/demo/resources. The command to restore the database from this dump file is:
mysql -u admin_user -padmin_password chanjo4_demo < path/to/chanjo4_structure.sql #no space between -p and passwordTo create instead a demo database containing coverage data for genes in panel1 for the samples used in the demo version of scout, you might download the schema chanjo4_demo.sql. This file is present in the folder scout/demo, along with the rest of the demo data. The syntax to create the database is the same as above.
Chanjo database consists of 3 tables:
- sample:
| Field | Type | Null | Key | Default | Extra |
|---|---|---|---|---|---|
| id | varchar(32) | NO | PRI | NULL | |
| group_id | varchar(128) | YES | MUL | NULL | |
| source | varchar(256) | YES | NULL | ||
| created_at | datetime | YES | NULL | ||
| name | varchar(128) | YES | NULL | ||
| group_name | varchar(128) | YES | NULL |
- transcript:
| Field | Type | Null | Key | Default | Extra |
|---|---|---|---|---|---|
| id | varchar(32) | NO | PRI | NULL | |
| gene_id | int(11) | NO | MUL | NULL | |
| gene_name | varchar(32) | YES | MUL | NULL | |
| chromosome | varchar(10) | YES | NULL | ||
| length | int(11) | YES | NULL |
- transcript_stat:
| Field | Type | Null | Key | Default | Extra |
|---|---|---|---|---|---|
| id | int(11) | NO | PRI | NULL | auto_increment |
| mean_coverage | float | NO | NULL | ||
| completeness_10 | float | YES | NULL | ||
| completeness_15 | float | YES | NULL | ||
| completeness_20 | float | YES | NULL | ||
| completeness_50 | float | YES | NULL | ||
| completeness_100 | float | YES | NULL | ||
| threshold | int(11) | YES | NULL | ||
| _incomplete_exons | text | YES | NULL | ||
| sample_id | varchar(32) | NO | MUL | NULL | |
| transcript_id | varchar(32) | NO | MUL | NULL |
This package might be downloaded and installed via git clone and using pip install command (instructions here), but perhaps the easiest way to make sure that it will serve pages once the app is invoked by scout is to install it with this command:
pip install scout-browser[coverage]Just make sure that both scout and chanjo-report are installed on the same virtual environment (if you are using one). Note that chanjo-report requires the package pymysql if it will be connecting to a MySQL database. So if it is not found in the environment you can install it with this command:
pip install pymysqlIn order to enable the support for the coverage report visualization in scout the configuration file named config.py under scout/server must be modified to include the following line:
SQLALCHEMY_DATABASE_URI = 'mysql+pymysql://username:password@host:port/chanjo4_demo'
The database URI provided in the line above refers to the MySQL database with the coverage data, so user and password are those used to connect to the database. Default database host and port are localhost and 3306 respectively.
If your scout implementation is using another configuration file other than the default config.py then the database connection URI must be added to this file instead.
Once all the above steps are executed you should see the "Coverage report" option on the left side bar under the scout case page. The buttons allow to generate HTML and PDF reports for the case samples.