About Speakers Schedule Transportation INS
The Workshop on Intelligent Computational Methods in Molecular Dynamics 分子动力学研究中的智能计算方法研讨会

Student's Report: Decoding Kinase Conformational Dynamics with Markov State Modelling

Speaker

Ryan Zhu , 爱丁堡大学 University of Edinburgh

Time

22 Nov, 16:40 - 16:50

Abstract

Protein kinases are essential regulators of cellular signaling. Dysregulation of kinase activity is a hallmark of many cancers, and their pharmaceutical importance is evidenced by over 80 FDA-approved inhibitors. Kinase inhibitors are known to target distinct binding sites and modulate the activity through different mechanisms. For example, type II inhibitors stabilize the inactive DFG-out conformation to suppress kinase activity. Despite the conserved overall fold, different protein kinases exhibit different conformational ensembles and transition pathways, reflecting their unique regulatory mechanisms. Therefore, understanding kinase dynamics and intermediate states within transition pathways provide mechanistic insights into their regulation and supports the rational design of selective inhibitors.

To compare the dynamic differences among kinases, we performed millisecond-scale parallel molecular dynamics simulations for three clinically relevant kinases Abl, EGFR, and c-Met, starting from both crystal and AlphaFold-predicted structures. We constructed Markov State Models (MSMs) from the trajectories to quantify their conformational landscapes and transition kinetics. The MSMs reveal that each kinase explores a unique sequence of intermediate states and transition routes between DFG-out and DFG-in conformations. Notably, we identified distinct bottleneck states that act as kinetic barriers, suggesting family-specific regulatory mechanisms. Together, these results provide a comparative view of kinase conformational landscapes, offering insights that can inform the rational design of inhibitors targeting transient or kinase-specific conformations.