Insights into four helical proteins folding via self-guided Langevin dynamics simulation

Abstract

Protein folding, vital to life and death, is the basis of biological physics. The information about the protein folding process and dynamics is of tremendous significance in the protein-folding field, and the research about protein folding pathways has attracted considerable attention. Here, four protein systems (1HOD, 2AJJ, 2DX3, 2RLG) are performed the folding simulation starting from the linear structures in order to investigate their folding mechanisms using the classical molecular dynamics (MD) and the self-guided Langevin dynamics (SGLD), respectively. The latter can increase the low-frequency motion, which is particularly related to the folding process of protein. Compared with MD, there is a significant improvement through SGLD method, such as the evolution of the radius of gyration, root mean square deviation, cluster analysis, native contact, helix content, process of the protein folding and the free energy landscape. This research shows that the extended states successfully fold into the corresponding native structures within 95 ns using SGLD. Furthermore, the result of multiple trajectories again supports the above conclusion. However, classical MD ultimately fails to show any stable helix formation at the same simulation time. The detailed folding pathway and mechanism of these proteins are elucidated using the SGLD method in this report.