Exploring Markov Processes Variational Approach for Dynamics Prediction of Antiviral Peptides

Resumo

Peptides are known to exhibit significant antiviral properties. This work focuses on the DENV 2 peptide (CGYGLC) in aqueous solution, notable for its anti-Dengue potential. Understanding its conformational dynamics is vital for assessing its structural and functional behavior. Molecular dynamics (MD) is commonly used for such studies, but is computationally intensive and time-consuming. To address this, we investigate the simulation of a reduced fraction of the total simulation time.

We employ the Variational Approach for Markov Processes (VAMP) to analyze the peptide’s conformational dynamics. The initial structure of DENV-2 was obtained using AlphaFold2 [2] and simulated for a total of 10 ns (equivalent to 50000 frames) with AMBER [1]. Dihedral angles define the polypeptide chain’s spatial conformation and are key to a peptide’s secondary and tertiary structure. These angles are derived by transforming Cartesian coordinates from molecular dynamics simulations. Torsion angles (ψ and ϕ) are favored because of their lower dimensionality, which is more closely related to the intrinsic dynamics of the system. Each dihedral angle is the torsion between four consecutive atomic positions. Their temporal evolution is modeled as a multivariate time series: [...]