inst/shiny-examples/myapp/tutorial.md

In gaelfortin/phosphogo: Analyze Your Phosphoproteomic Data

What is phosphogo?

Phosphogo is a set of tools to analyze phosphoproteomic data from mouse and human experiments. Kinase-substrate predictions can be performed using NetworKIN and IV-KEA (in vitro kinase enrichment analysis) in a user-friendly interface. Analyzing phosphoproteomic data without knowing how to code is now posible thanks to Phosphogo!

Phosphogo user interface

This user interface allows you to run the same tools than the command-line version of phosphogo. It requires the package shiny. From this application, you can install or reinstall phosphogo dependencies, input your phosphoproteomic data, generate predictions and visualize outputs.

0. Setup

You can install/reinstall phosphogo dependencies on this page. Phosphogo scans the folder from which the app is launched and will install NetworKIN and/or the IV-KEA database in this folder. NetworKIN and the IV-KEA database will be downloaded and installed in the current folder.

1. Your input

Import your data

This page allows you to import your data as a .csv or .xlsx table.

Indicate the colums to use

The table should contain the following columns: - The name of each phosphosite as UniprotID-Phospho:Site (e.g PCDH7-Phospho:S1011) - Log2 of expression ratio between control and condition. Recommended formula is log2(condition/control). - Adjusted p-value

The section Your data at a glance will output the first lines of the uploaded table to help you select the right columns.

Select the species

Indicate the species of origin of your data. Mouse data will be converted into human data for NetworKIN and IV-KEA using the Phosphosite orthology database.

Indicate where outputs should be saved

Finally, select where to output phosphogo files. These files include pre-processing files, raw prediction outputs, graphs... Plots can still be generated using the 3. Visualize tab of phosphogo.

2. Run

You can run either NetworKIN or IV-KEA.

IV-KEA

In vitro kinase enrichment analysis relies on an in vitro database. This analysis is fast and relies on in vitro data but is restricted to 300 kinases. Find more details in Sugiyama et al., 2019.

NetworKIN

NetworKIN is a predictive algorithm that uses published datasets and homology to predict kinase-subtrate relationship. For more details, see Horn et al., 2014.

Quality control

Phosphogo selects by default only the top 20% predictions according to their prediction score. Phosphogo will output the density of prediction scores as a quality control.

NetworKIN can output hundreds of predictions per phosphosite, often with a low prediction score. Thus, a high density of low scores is expected in usual conditions.

3. Visualize

You can visualize the results of kinase enrichment analysis. These results reveal kinases that are differentially predicted more/less often in your condition compared to control. The generated graphs are interactive, their range and zoom can be changed and they can be saved thanks to the tools at the top right side of the graphs.

Enrichment graph

The results of one enrichment analysis (either IV-KEA or NetworKIN) can be visualized in the form of an interactive volcano plot. The enrichment results files are outputted by default by phosphogo and start by kinase_enrichment (usually kinase_enrichment_ivkea.csv or kinase_enrichment_networkin.csv).

Comparison graph

If you have run NetworKIN and IV-KEa on your dataset, you can easily compare the results of both algorithms in one plot thanks to this function. Just import the two kinase enrichment files and generate the graph.

As NetworKIN and IV-KEA use two different annotation systems, some kinases are labelled with different names by the to algorithms.

gaelfortin/phosphogo documentation built on Nov. 26, 2020, 9:12 a.m.