Erik Koshnicharski
Binghamton University
Combining the Martini Model and Gō potentials for Peptide Folding
Peptides are short chains of amino acids with an increasingly important role in biomaterials and drug design. They can be difficult to probe experimentally because they often exist in equilibrium between unfolded, disordered states and folded, ordered states. As such, computational techniques such as molecular dynamics (MD) simulations are a good complement to ensemble-based experimental results. Martini3 is a widely used coarse-grained force field; however, it fails to capture the equilibrium of folded and unfolded states because its formulation restricts the peptide to the predefined secondary structure of the input. The prescriptive interactions can take the form of strong dihedrals or harmonic (elastic network) potentials. Here, we propose weakening the backbone dihedrals and adding a Gō potential to retain folded states and allow exploration of unfolded states. The new layer of Gō potentials can provide a tunable barrier between folded and unfolded states. Using a diverse set of 20 peptides, we analyze the folding free energy landscape, peptide root mean square deviation, and fraction of native contacts time series. Together, we show that the Martini3-IDP-Gō model is able to tunably capture the conformational flexibility of peptides.
Peptides are short chains of amino acids with an increasingly important role in biomaterials and drug design. They can be difficult to probe experimentally because they often exist in equilibrium between unfolded, disordered states and folded, ordered states. As such, computational techniques such as molecular dynamics (MD) simulations are a good complement to ensemble-based experimental results. Martini3 is a widely used coarse-grained force field; however, it fails to capture the equilibrium of folded and unfolded states because its formulation restricts the peptide to the predefined secondary structure of the input. The prescriptive interactions can take the form of strong dihedrals or harmonic (elastic network) potentials. Here, we propose weakening the backbone dihedrals and adding a Gō potential to retain folded states and allow exploration of unfolded states. The new layer of Gō potentials can provide a tunable barrier between folded and unfolded states. Using a diverse set of 20 peptides, we analyze the folding free energy landscape, peptide root mean square deviation, and fraction of native contacts time series. Together, we show that the Martini3-IDP-Gō model is able to tunably capture the conformational flexibility of peptides.
