For beginners

The interactive interface

We designed an interactive interface (a viewer) for you if you are not famaliar with programming. Here is a tutorial.

Before using the viewer, please install iMVP-utils first (see Installation), then run the script iMVP_viewer.py by typing it in your shell:

iMVP_viewer.py

this operation will open a backend of the iMVP viewer, and you can visit the application via your browser. By default, please type 127.0.0.1:8050 in your URL box to access it.

If you want to run the backend on your sever but work with it in another computer, you can assign a IP address for it:

iMVP_viewer.py --host xxx:xxx:xxx:xxx

here, xxx:xxx:xxx:xxx should be the IP address of the sever, for example 123.123.123.123. Then you can visit from the browser of another computer with xxx:xxx:xxx:xxx:8050. Use --port option if you want to switch to another port.

You will visit this page when you succeed in launching the application. You can drag your FASTA file to the upload box, or just click it to send your file to the sever. Then you should adjust the parameters on the right. When everything is ready, click “submit” to run iMVP

_images/Guide_1.png

When UMAP and HDBSCAN finished, you will find a “Draw the figure” button, click it to get the figure.

../Images/Guide/Guide_2.png

And you will find your UMAP projections with the clusters labled in different colors. You can click the check box to select one or several groups classified by HDBSCAN. You can download the FASTA file and the PNG logo file by clicking the corresponding button.

_images/Guide_3.png

Of course, you can also use a selector/lasso to plot a set of sites manually.

_images/Guide_4.png

Tip

Double click the figure to release the selector.

If you want to modify your parameters, please refresh the page.

Use Ctrl + C to terminate the backend.

Simple notebook examples

To make it clear, we use Jupyter-notebook (https://jupyter.org/) to present the backend of iMVP. With this tutorial, you can easily set up your Jupyter-notebook server in your computer. When you set up Jupyter-notebook, you can simply repeat our scripts.

You can work with the well packaged functions in beginners.ipynb.

After clustering

We suggest you to plot the clusters with Weblogo (https://weblogo.berkeley.edu/) directly. You can also analyze the sequences from the clusters with MEME (https://meme-suite.org/meme/doc/meme.html), HOMER (http://homer.ucsd.edu/homer/), and other motif finders.