Skip to content

Comparison of different methods for kinase activity estimation

Notifications You must be signed in to change notification settings

saezlab/kinase_benchmark

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Systematic comparison of methods for kinase activity estimation

Overview

In this study, we present a flexible framework to assess different combinations of computational algorithms and kinase-substrate libraries for the inference of kinase activities. For the benchmark, we use a set of kinase perturbation experiments to evaluate which combination is able to recapitulate the perturbed kinases from the phosphoproteomics data.

If you want to test your own method try out our package benchmarKIN and check out the documentation.

Kinase substrate libraries

We have included the following kinase-substrate libraries:

  • PhosphoSitePlus
  • PTMsigDB
  • Omnipath
  • Gold Standard set of GPS 6.0
  • iKiP-db
  • NetworKIN

Additionally have tested the combination with predicted targets including the Kinase Library and Phosformer.

Methods

We have included the following methods for the comparison:

  • fgsea
  • KARP
  • Kologomorov-Smirnov
  • linear model (RoKAI, decoupler)
  • mean
  • normalised mean
  • median
  • multivarite linear model
  • PC1
  • PTM-SEA
  • sum
  • upper quantile
  • viper
  • Wilcox
  • z-score (KSEA, RoKAI)

Citation

Mueller-Dott, Sophia, Eric J. Jaehnig, Khoi Pham Munchic, Wen Jiang, Tomer M. Yaron-Barir, Sara R. Savage, Martin Garrido-Rodriguez, et al. 2024. “Comprehensive Evaluation of Phosphoproteomic-Based Kinase Activity Inference.” bioRxiv. https://doi.org/10.1101/2024.06.27.601117.

About

Comparison of different methods for kinase activity estimation

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages