Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
61 changes: 61 additions & 0 deletions _posts/2025-10-23-release-v1.7.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,61 @@
---
layout: post
title: "OpenFE v1.7: SepTop, ABFEs, faster simulations, and more!"
categories: science update
---

We are pleased to announce the release of **openfe** version 1.7.0!

This release brings along some long-awaited features, notably including new Protocols (SepTop & ABFEs), and performance improvements.

Please see below for some highlights and see the [Changelog][] for more details.

## New alchemical simulation protocols

### Separated Topologies

We include a new relative binding free energy protocol based on the Separated Topologies (SepTop) method, as defined by [Rocklin et al.][] and [Baumann et al.][].
Using SepTop, you can calculate the difference in binding free energy between two ligands by essentially performing two absolute binding free energy calculations simultaneously in opposite directions.

Crucially, this avoids the needs for creating a hybrid topology, **making atom mapping unnecessary and allowing for scaffold hopping!**

Please see our [SepTop tutorial][] for a demonstration on how to use it! You can also read more about the method in our [SepTop user guide][] entry.

### Absolute Binding Free Energies

We also include a new Absolute Binding Free Energy (ABFE) protocol, based on a partial annhilation scheme. This method allows for the direct calculation of binding affinities by alchemically removing a ligand from its binding site.

Please see our [ABFE tutorial][] and associcated [ABFE user guide entry][] to see how to use the method.

## Performance improvements through better default settings

We have updated our default settings to maximize the performance of `openfe`, leading to a **~2x performance improvement compared to the v1.0 release!**

The figure below shows the average performance of a single thermodynamic cycle for a pairwise transformation using our relative binding free energy hybrid topology protocol. We see see a drop from ~6 hours to ~3 hours by switching our default settings and moving to OpenMM 8.2. This can be further improved by layering multiple simulation using the [NVIDIA multi-processing service][]

<center><img src="/assets/images/per_benchmark_presentation.png" alt="average cost of a ligand pair alchemical cycle over the JACS set." style="width:550px;"/></center>

These settings updates mostly involve moving away from overly conservative parameters to ones that are more widely used in production settings. These include:
* Switching from a cubic box to a dodecahedron box for solvation, avoiding excess waters.
* Increasing the replica exchange rate to 2.5 ps (from 1 ps), reducing performance losses linked to GPU context switches.
* Reducing the nonbonded cutoff from 1.0 nm to the OpenFF recommended value of 0.9 nm.

These new default settings were extensively validated on RBFE and AHFE simulations, showing little to no difference against the previous defaults.

## Installing `openfe v1.7.0`

`openfe` is made available as a [conda-forge][] package, or [docker and singularity images][].

Please see our [installation instructions][] for details on how to install `openfe`.

[Rocklin et al.]: https://doi.org/10.1063/1.4792251
[Baumann et al.]: https://doi.org/10.1021/acs.jctc.3c00282
[SepTop tutorial]: https://docs.openfree.energy/en/latest/tutorials/septop_tutorial.html
[SepTop user guide]: https://docs.openfree.energy/en/latest/guide/protocols/septop.html
[ABFE tutorial]: https://docs.openfree.energy/en/latest/tutorials/abfe_tutorial.html
[ABFE user guide entry]: https://docs.openfree.energy/en/latest/guide/protocols/absolutebinding.html
[Changelog]: https://docs.openfree.energy/en/latest/CHANGELOG.html
[NVIDIA multi-processing service]: https://developer.nvidia.com/blog/maximizing-openmm-molecular-dynamics-throughput-with-nvidia-multi-process-service/
[docker and singularity images]: https://github.com/OpenFreeEnergy/openfe/pkgs/container/openfe
[conda-forge]: https://anaconda.org/conda-forge/openfe
[installation instructions]: https://docs.openfree.energy/en/latest/installation.html
Binary file added assets/images/per_benchmark_presentation.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading