Re: [GaMD-discuss] Peptide Gaussian Accelerated MD - Question about Simulation Set-up

[Miao, Y. <Yin...@me...>]

Hi Arpan,

Thanks for your interest in Pep-GaMD!

Yes, it would be great if you can share your simulation input parameter files and probably mdout file for the Pep-GaMD equilibration. We can then look into the issue in more detail.

Best regards,
Yinglong

Yinglong Miao, Ph.D.
Associate Professor
Department of Pharmacology
Computational Medicine Program
Lineberger Comprehensive Cancer Center
University of North Carolina - Chapel Hill
Tel: 1-919-962-5696
http://miaolab.org<http://miaolab.org/>

Editor-in-Chief
npj Drug Discovery
https://www.nature.com/npjdrugdiscov

On Jan 30, 2026, at 11:02 AM, Arpan Tapdiya <agt...@nc...> wrote:


Hello Prof Miao,

My name is Arpan Tapdiya, and I am a PhD student in Professor Carol Hall’s research group in the Department of Chemical and Biomolecular Engineering at North Carolina State University. Our group designs short peptides for biosensor and therapeutic applications. To do so, we have developed our Monte-Carlo based Peptide Binding Design Algorithm (PepBD), and we use molecular dynamics (MD) and enhanced sampling to quantify peptide-target interactions.

I am currently applying Pep-GaMD to study binding dynamics for our designed 10-residue peptides interacting with a ~240-residue protein target. I set up simulations following the Pep-GaMD protocol from your paper<https://doi.org/10.1063/5.0021399> (with a larger solvent box for this system) and explored different boost settings, including varying σ0P and σ0D from ~2 to 6 kcal/mol.

While the peptide behaves as expected by dissociating and exploring bound/unbound regions, during the GaMD equilibration stage (when the boost is applied to the target protein + solvent in the dual-boost scheme), the target protein begins to lose its native secondary structure. The target has four α-helices at the start of the simulation, but shortly after the Gaussian boost is applied, it progressively unfolds and appears intrinsically disordered. This occurs even at relatively small boost settings. Notably, the system remains stable in conventional MD without the GaMD boost.

The target structure is taken from the Protein Data Bank, and the peptide was designed through our simulation workflow.

I would be very grateful for your guidance on the following:

  1.  What are the most common causes of target protein unfolding in Pep-GaMD/GaMD runs?

  2.  In your experience, can unfolding indicate that the boost potential is effectively too aggressive for the system even when σ0P/σ0D are set low? If so, what parameter adjustments or protocol changes would you recommend?

  3.  Are there recommended strategies to maintain target stability in Pep-GaMD without adding additional constraints, while still allowing meaningful peptide binding/unbinding sampling?

If helpful, I can share the simulation setup details and the necessary files.

Thank you very much for your time and for developing and sharing GaMD/Pep-GaMD. This method is extremely valuable for the types of peptide systems we study, and we will be thankful for any help!

Sincerely,
Arpan

--
Arpan Tapdiya
Graduate Student
Department of Chemical and Biomolecular Engineering
North Carolina State University