An optimal set of inhibitors for Reverse Engineering via Kinase Regularization

Abstract We present a comprehensive resource of 257 kinase inhibitor profiles against 365 human protein kinases using gold-standard kinase activity assays. We show the utility of this dataset with an improved version of Kinome Regularization (KiR) to deconvolve protein kinases involved in a cellular...

Full description

Saved in:
Bibliographic Details
Published inbioRxiv
Main Authors Rata, Scott, Gruver, Jonathan Scott, Trikoz, Natalia, Lukyanov, Alexander, Vultaggio, Janelle, Ceribelli, Michele, Craig, Thomas, Taran Singh Gujral, Kirschner, Marc W, Peshkin, Leonid
Format Paper
LanguageEnglish
Published Cold Spring Harbor Cold Spring Harbor Laboratory Press 28.09.2020
Subjects
Cell migration
Kinases
Learning algorithms
Machine learning
Phenotypes
Protein kinase inhibitors
Proteins
Online AccessGet full text

Cover

More Information
Summary:Abstract We present a comprehensive resource of 257 kinase inhibitor profiles against 365 human protein kinases using gold-standard kinase activity assays. We show the utility of this dataset with an improved version of Kinome Regularization (KiR) to deconvolve protein kinases involved in a cellular phenotype. We assayed protein kinase inhibitors against more than 70% of the human protein kinome and chose an optimal subset of 58 inhibitors to assay at ten doses across four orders of magnitude. We demonstrate the effectiveness of KiR to identify key kinases by using a quantitative cell migration assay and updated machine learning methods. This approach can be widely applied to biological problems for which a quantitative phenotype can be measured and which can be perturbed with our set of kinase inhibitors. Competing Interest Statement The authors have declared no competing interest.
DOI:10.1101/2020.09.26.312348