One workflow, many starting points: a practical guide to molecular simulation batches

Running the same molecular-dynamics workflow repeatedly can turn a straightforward modeling task into a setup chore. You may need to test many conformations of one molecule, or apply consistent simulation settings to several related systems. Preparing each project by hand makes it harder to keep the workflow consistent and easier to lose track of which result came from which starting state.

GROMACS Wizard’s batch workflow in SAMSON helps organize these jobs. The key is to decide which kind of batch you have before you begin: one system with many conformations, or several systems that have each been prepared separately. The distinction affects how you set up the batch and, for conformations, how you choose the periodic box.

When the system is the same but the starting state changes

Use this route when you want to run the same workflow from several conformations of one molecular system. Starting states can come from a trajectory or from a set of conformations already in SAMSON. If you have a trajectory, open it in SAMSON or drag it into the application. You can use the whole path, where each frame becomes a starting state, or select individual conformations instead.

There is no need to use every trajectory frame if that is not useful for your study. You can inspect frames and create conformations for selected ones. Alternatively, create conformations from the path and remove the ones you do not want to include. This gives you a deliberate set of starting states rather than an unnecessarily large batch.

In the GROMACS Wizard’s Prepare tab, set the molecular system, enable [Optional] Batch preparation, select the conformations or path in the Document view, and choose Set conformations or a path. Then follow the usual preparation choices for the force field, solvent, and ions. The batch project will contain numbered subfolders, each representing a separate project.

The box-size decision: consistency or compactness

For a conformational batch, think about whether every subproject needs the same box size. With Box lengths, the initial box is based on all the conformations or the whole path, so the batch shares a box size. This can make the setup consistent across starting states, but the box must accommodate the system’s extent across them. A tightly fitted box may need to be enlarged to satisfy the minimum image convention.

With Solute-box distance, the box is sized separately for each conformation. The documentation recommends a distance of at least 1 nm to help ensure the minimum image convention. This option can give compact conformations smaller boxes, which can reduce computational time, but box sizes will differ across the batch. Use the preparation slider to inspect how the selected box fits different conformations.

A useful rule of thumb: choose shared box lengths when matching box dimensions across states matters; choose per-conformation solute-box distance when different box dimensions are acceptable and smaller boxes for compact states are useful.

When the batch contains different systems

For multiple systems, prepare each one separately first. Put the resulting folders ending in -prepare together inside one batch folder. You can rename the subfolders, for example with PDB identifiers; those names are preserved in later steps. During minimization or another subsequent step, select From batch folder and point the input to that folder.

One important compatibility check: avoid MD parameters that refer to index groups or position restraints (POSRES records) present in only some batch members. A setting that works for one system may not apply to the others. This is especially relevant when using advanced options for complex systems.

Continue the workflow as a batch

After preparation, the next stages follow the familiar single-project sequence: energy minimization, NVT equilibration, NPT equilibration, and production molecular dynamics. For each stage, choose From batch folder as the input. The auto-fill button can set the input path based on the previous run.

Selecting a batch project as the input source

Batch jobs can run locally or in the Cloud. For local runs, a separate job is created for each subfolder; the Local jobs window lets you monitor and manage those jobs. Cloud runs likewise create a separate Cloud job per subfolder, which you can monitor, stop or cancel, download, and import.

Before launching, check that the batch represents the intended starting states or systems, that the box choice suits your comparison, and that shared parameters are valid for every member. Those small checks help make a batch easier to interpret later.

For the full walkthrough, see the GROMACS Wizard batch computations tutorial.

Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.

Comments are closed.