Uncovering phosphorylation–based specificities through functional interaction networks
Select a prediction from the dropdown below to get started
About
The method presented here aims at predicting specificities of kinases. Although the method can be expanded to other enzymes that target post translational modifications.
The idea behind this method is rather simple. The assumption here is that putative interaction partners of a kinase are more likely than random proteins to be phosphorylated by that kinase. Therefore, phosphosites occurring on interaction partners of kinases should confer a bias in amino acid composition towards the kinase’s specificity. By performing motif enrichment on these phosphosites, the specificity of the kinase can, in some cases be predicted (see figure below).
Acknowledgements
We gratefully thank the following resources for providing phosphorylation sites and kinase substrates
Contact
If you have any questions, comments or suggestions get in touch with us.
Tutorial
A tutorial on how to generate the predictions yourself in R can be found here
Tips and tricks
- Use the ← or → keys to navigate through pages of the sites table
- Click the logo images for a zoomed in version
- Hit the toggle button to switch between bits and probability views for the logo
- Filter the sites table by typing into the search bar on the top right
Download data
For convenience, links to download all predictions and logos are provided below. In addition to these we provide other datasets including all protein sequences, STRING network (v9.1) and all phosphosites used.
All TSV flatfiles (zip) All PDF bit logos (zip) All PDF probability logos (zip) All other data (zip)Please note that predicted models are speculative and should be used with caution. They are based on motifs extracted from phosphosites belonging to the kinase's putative interaction partners. We advise users who use this dataset to manually verify that the logos appear valid before proceeding with downstream analyses.
[]
[]
Details
- - Group/family
- Yes Proline +1 sequences removed
- 55 Number of interacting partners
- 38 Number of interacting partners with at least one site
- 366 Number of sites on partners
Motifs
| Motif | Score | FG matches | FG size | BG matches | BG size | Fold increase |
|---|
Phosphosites
| Protein | String score | Sequence | Position | Kinase | Source |
|---|