kmer-profile

A Dash app for interactive visualization of read profiles generated with FASTK.

NOTE: This package also contains python version of ClassPro, but it is only for development and not intended for usage by users.

Requirements

Followings are required to be installed before installing this package (i.e. these are not automatically installed):

• Python >= 3.7
• FASTK
• FASTK_python

Optional requirements

• Followings must be installed in your PATH if necessary
  • Seqkit (if you use fasta/fastq files)
  • DAZZ_DB (if you use db/dam files)

How to install

$ git clone https://github.com/yoshihikosuzuki/kmer-profile
$ cd kmer-profile
$ python3 setup.py install

A binary executable for visualization, kmer_profile, is installed in addition to a python module.

How to use the visualizer

You are supposed to have FastK's outputs with the -p option (i.e. you need to have a .prof file) for your dataset.

You can start the visualizer with the command named kmer_profiler. The usage is as follows (this message can also be shown via $ kmer_profile -h):

usage: kmer_profiler [-h] [-s SEQ_FNAME] [-c CLASS_FNAME]
                     [-t TRUTH_CLASS_FNAME] [-p PORT_NUMBER] [-f IMG_FORMAT]
                     [-d]
                     fastk_prefix

K-mer count profile visualizer

positional arguments:
  fastk_prefix          Prefix of FastK's output files. Both
                        `<fastk_prefix>.hist` and `<fastk_prefix>.prof` must
                        exist.

optional arguments:
  -h, --help            show this help message and exit
  -s SEQ_FNAME, --seq_fname SEQ_FNAME
                        Name of the input file for FastK containing reads.
                        Must be .db/dam/fast[a|q]. Used for displaying baes in
                        profile plot [None]
  -c CLASS_FNAME, --class_fname CLASS_FNAME
                        File name of K-mer classification result. [None]
  -t TRUTH_CLASS_FNAME, --truth_class_fname TRUTH_CLASS_FNAME
                        File name of ground truth of K-mer classification
                        result. [None]
  -p PORT_NUMBER, --port_number PORT_NUMBER
                        Port number of localhost to run the server. [8050]
  -f IMG_FORMAT, --img_format IMG_FORMAT
                        Format of plot images you can download with camera
                        icon. ['svg']
  -d, --debug_mode      Run a Dash server in a debug mode.

After running this, open http://localhost:8050 (port number can be changed with the option -p) in a browser.

[NOTE] To use DOWNLOAD HTML buttons in the app, you might need to disable your browser's cache (only of the appr's tab). To do this, e.g. in Chrome, go to Developer Tool → Network tab → Disable cache checkbox and keep the Developer Tool open while using the viewer.