A structural machine learning approach for rapid prediction of thermodynamically destabilizing tyrosine phosphorylations

Jaie Woodard; Zhengqing Liu; Atena Malemir Chegini; Jian Tian; Rupa Bhowmick; Subramaniam Pennathur; Alireza Mashaghi; Jeffrey R Brender; Sriram Chandrasekaran

doi:10.1016/j.crmeth.2025.101169

A structural machine learning approach for rapid prediction of thermodynamically destabilizing tyrosine phosphorylations

Cell Rep Methods. 2025 Sep 15;5(9):101169. doi: 10.1016/j.crmeth.2025.101169.

Authors

Jaie Woodard¹, Zhengqing Liu¹, Atena Malemir Chegini², Jian Tian³, Rupa Bhowmick¹, Subramaniam Pennathur⁴, Alireza Mashaghi², Jeffrey R Brender⁵, Sriram Chandrasekaran⁶

Affiliations

¹ Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI 48109, USA.
² Medical Systems Biophysics and Bioengineering, Leiden Academic Centre for Drug Research, Faculty of Science, Leiden University, 2333CC Leiden, the Netherlands.
³ State Key Laboratory of Animal Nutrition and Feeding, Institute of Animal Sciences, Chinese Academy of Agricultural Sciences, Beijing 100193, China.
⁴ Department of Medicine and Molecular and Integrative Physiology, University of Michigan, Ann Arbor, MI 48109, USA.
⁵ Molecular Imaging Program, National Institutes of Health, Bethesda, MD 20892, USA; Clinical Cancer Metabolism Facility, Center for Cancer Research, National Cancer Institute, Bethesda, MD, USA.
⁶ Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI 48109, USA; Center for Computational Medicine and Bioinformatics, University of Michigan, Ann Arbor, MI 48109, USA; Rogel Cancer Center, University of Michigan, Ann Arbor, MI 48109, USA. Electronic address: csriram@umich.edu.

Abstract

Tyrosine phosphorylations are a prominent characteristic of numerous diseases, yet it is challenging to identify potentially (dys)functional phosphorylations among thousands of phospho-proteins. Here, we propose a machine learning method to predict the thermodynamic stability change resulting from tyrosine phosphorylation. Our approach, based on the prediction of phosphomimetic stability (ΔΔG) from structural features, strongly correlates with experimental phosphorylation stability and mutational scanning cDNA proteolysis data (R = 0.55-0.67). We apply our approach to predict the potential destabilizing effects of all 384,858 tyrosine residues from the Alphafold2 database, the PhosphoSitePlus database, and on a pan-cancer phosphoproteomics dataset with 11 cancer subtypes. We predict destabilizing phosphorylations in both oncogenes and tumor suppressors, and ΔΔG values and local protein circuit topology features are able to distinguish phospho-proteins that are known to be dysregulated in cancer. Our approach can enable rapid screening of destabilizing phosphorylations and phosphomimetic mutations.

Keywords: CP: computational biology; computational biophysics; phosphorylation; post-translational modifications; protein stability; systems biology.

MeSH terms

Humans
Machine Learning*
Mutation
Neoplasms / genetics
Neoplasms / metabolism
Phosphoproteins / chemistry
Phosphoproteins / metabolism
Phosphorylation
Thermodynamics
Tyrosine* / chemistry
Tyrosine* / metabolism

Substances

Tyrosine
Phosphoproteins