Installation

Environment

Most of the functions in iMVP-utils can be run on Linux, Windows, and MacOS.

Our interactive interface is not available on Windows because some applications (e.g., Weblogo) are not supported. For such a situtation, we suggest you to use WSL (https://docs.microsoft.com/en-us/windows/wsl/).

Dependencies

The workflow is supported by Python 3.

Tested/recommended versions are indicated. In each notebook, we also highlight the required packages.

Basic Requirements

umap-learn (>=0.5.2)
hdbscan (>=0.8.27)
pandas (>=1.3.4)
numpy (>=1.20.0)
scipy (>=1.5.1)
scikit-learn (>=0.23.1)
biopython (>=1.77)

Notebooks

jupyter-notebook

Approximate clustering

cv2 (>=4.5.5)

Figures generation

matplotlib (>=3.2.2)
seaborn (>=0.10.1)
weblogo (>=3.7.0)

Interactive interface

dash (>=2.0.0)
dash-bio (>=0.9.0)
imageio (>=2.13.5)

GPU accerleration

cuML

See GPU_accerleration for details.

Install from Pypi

To install the package with pip:

pip install iMVP_utils

Install in a virtual environment

We recommend to use pipenv to avoid interrupting your environment variables. If pipenv is not istalled, run:

pip instasll pipenv

To install the package in a virtual environment (Python3 only):

# make a directory for iMVP
mkdir iMVP
cd iMVP
pipenv install -r requirements.txt
pipenv install iMVP_utils

Then run

pipenv shell
# run iMVP-utils in an interactive mode
iMVP_utils

To exit the environment:

exit

Successful installation

To check if iMVP-utils and the interative interface are installed correctly. You can type

iMVP_viewer.py --help

in your shell, then you will see the help message:

usage: iMVP_viewer [-h] [--output OUTPUT_PATH] [--host HOST] [--port PORT]

optional arguments:
-h, --help            show this help message and exit
--output OUTPUT_PATH, -o OUTPUT_PATH
                        The output directory. If not exist, will make the folder. Default="./output/"
--host HOST           The IP address of the app, default=127.0.0.1 (localhost)
--port PORT           The port number that you want to dispaly the app.

You can also enter the Python shell to load the package:

import iMVP_utils