Quick Start

Using ngreport

High level NGCloud report generator can be accessed by both ways,

$ ngreport
# Usage:
#     ngreport [-p <pipeline>] <job_dir> [<out_dir>] [-v ...]
#     ngreport [-p <pipeline>] <job_dir> [-o <out_dir>] [-v ...]
#     ngreport -h | --help
#     ngreport --version

$ python -m ngcloud.report
# Usage:
# ... (they access the same function)

See also

To use the parser inside a Python program, call ngcloud.report.gen_report() directly.

It takes three main arguments to run:

  • Valid pipeline name (default: tuxedo)
  • Path to job folder (required)
  • Path to output folder (default: ./output)

Currently, only Tuxedo pipeline is valid, so it is the default option. You can still pass -p tuxedo or -p ngcloud.pipe.tuxedo.TuxedoReport to tell ngreport to generate the report of Tuxedo pipeline.

If you want to use your own NGS results running through the same pipeline, you need to organize the result folder and write some extra information so NGCloud can recognize. See Prepare Your NGS Result for NGCloud for more information.

By now, just follow our steps and use our bundled minimal example. We shipped a minimal example with NGCloud source code, hence you can use it to give a first try. Finally, you need to decide the output location. We choose /tmp for demo. By default it is ./output.

A full help message can be get by ngreport -h.

Run the minimal example

To get familiar with how NGCloud works, we first demo the usage through a minimal Tuxedo NGS pipeline example. Note that the example is for demo usage, don’t use the generated report for research.

The minimal result is under ngcloud_src/examples/job_tuxedo_minimal.

To generate the report, run

ngreport -p tuxedo {ngcloud_src}/examples/job_tuxedo_minimal -o /tmp

You need to replace {ngcloud_src} with the path where you download the NGCloud source code. If no warning message is shown, the output will be under /tmp/report_9527. It is a HTML-based report so you need a web browser to view it. Open /tmp/report_9527/index.html and you can get the beautiful report generated by NGCloud.

../_images/minimal_example_screenshot.png

Screenshot of the report frontpage

Click on other pages such as Quality Control to quick peek the report. Now you have finish the first run using NGCloud. Congrats!

If you want to see what’s going on during the report generation, pass an extra -v argument to ngreport to increase the verbosity. A colorful log output will be produced by extra --color argument, which requires colorlog or install NGCloud with pip install ngcloud[color].

Further Reading

On the next page, we will show you how to organize your own NGS result to feed in NGCloud. Also, a detailed explanation will be given to show what’s under the hood.

If your NGS analysis pipeline is not supported by NGCloud, you can add it your own, combining the existed NGCloud report templates. This tutorial will show you how to write your own pipeline. You will need NGCloud API documentation for more information.

Finally, If you enjoy using NGCloud and hope it gets better, the best support will be contributing back to NGCloud. See Contributing to get involved.