open force field

An open and collaborative approach to better force fields

Webinars
Workshops
Event calendar

Workshops

Please see below the list of the past and upcoming workshops organized by the Open Force Field Initiative.

3rd Open Force Field Workshop

The Third Open Force Field Workshop was initially as an in-person gathering in Boston on May 4-5, 2020, but this meeting was cancelled due to the Covid-19 outbreak. This in-person meeting was replaced by a virtual meeting on the same dates. We decided to slightly experiment with the workshop format by making most of the content available before the meeting for asynchronous consumption.

Read more
Read less

May 4-5, 2020 | Virtual meeting

We hosted the Third Open Force Field Consortium Workshop on May 4-5, 2020 as a virtual meeting. This virtual meeting replaced the planned in-person gathering in Boston. All talks were pre-recorded and the meeting time was focused on discussions and future planning. David Cerutti, who has joined the Open Force Field Initiative as a senior scientist to assist with development of biopolymer force fields funded by the NIH grant, gave a virtual keynote and a brief summary of his talk before discussion. The main aim of this format was to make this virtual meeting more interactive, and hopefully, more productive.

Talks and discussions

The recording sessions were open to all meeting attendees to join in real time, and shared on our YouTube channel on the same day. To access presentation slides and better video quality, download provided content from Zenodo. All talks with links to YouTube and Zenodo are provided below together with meeting notes from discussion sessions.


Agenda

Talks

The list of talks grouped by theme:

Introduction

David Mobley: Current status of OpenFF and our near-term roadmap
YouTube | Zenodo

Infrastructure Update

Daniel Smith: QC Archive infrastructure updates
YouTube | Zenodo
Jeffrey Wagner: Updates on core OpenFF infrastructure
YouTube | Zenodo
Chaya Stern: Capturing non-local through-bond effects when fragmenting molecules for QC torsion scans
YouTube | Zenodo

Force Field Updates

Jessica Maat: Training dataset selection
YouTube | Zenodo
Hyesu Jang: Update on Parsley minor releases (openff-1.1.0, 1.2.0)
YouTube | Zenodo

Benchmarking (Parsley)

Victoria Lim: Benchmark assessment of molecular geometries and energies from small molecule force fields
YouTube | Zenodo
David Hahn: Parsley parameters in protein-ligand binding free energy calculations
YouTube | Zenodo

Next Steps / Ongoing Studies

Simon Boothroyd: Which physical properties should we be optimizing against?
YouTube | Zenodo
Jeffry Setiadi: Host-guest binding calculations with OpenFF evaluator
YouTube | Zenodo
Michael Gilson: How Many Lennard-Jones Atom Types Does a Force Field Need?
YouTube | Zenodo

Future Directions

Owen Madin: Future directions in parameterization: enabling Bayesian inference with surrogate modeling
YouTube | Zenodo
Yuanqing Wang: Near-chemical accuracy at molecular mechanics speed: Neural parameterization of extended force field using graph nets
YouTube | Zenodo
Yuanqing Wang: Graph Nets for partial charge prediction
This is a previously given webinar, but relevant for this context.
YouTube | Zenodo

Biopolymers / Keynote

David Cerutti: Strategies for ab initio Biomolecular Force Field Development
YouTube | Zenodo


Discussion sessions (May 4-5, 2020)

Please note that all times refer to the Pacific Time Zone.

Day 1 (May 4, 2020)

Meeting notes - Day 1

Time Discussion
7.00 am Session I: Current status and performance of Open Force Fields and infrastructure
- David Mobley: Introduction and a brief recap from talks
- Discussion
Relevant talks (by theme): Introduction, Infrastructure, Force field updates, Benchmarking
8.30 am Break
8.45 am Session II: Roadmaps
- Science
- Infrastructure
- Data
Relevant talks (by theme): Infrastructure, Force Field Updates, Next Steps

Day 2 (May 5, 2020)

Meeting notes - Day 2

Time Discussion
7.00 am Session I: Biopolymer force fields
- Michael Shirts: Aims for the NIH-funded bioplymer force fields
- David Cerutti: Strategies for ab initio biomolecular force field development
YouTube | Zenodo
- Discussion
8.30 am Break
8.45 am Session II: Future directions
- John Chodera: Future directions for the Open Force Field Initiative
Relevant talks (by theme): Future Directions
8.30 am Wrap-up discussion

Recording sessions

Please note that all times refer to the Pacific Time Zone.

Day 1 (Apr 27, 2020)

Time Speaker Talk
8.30 am Chaya Stern Capturing non-local through-bond effects when fragmenting molecules for QC torsion scans
Wiberg bond orders (WBO) are sensitive to remote substituents and can be used as a surrogate for torsion energy barrier heights. Here we use WBOs to assess the robustness of fragmentation schemes and identify conjugated bond sets.
YouTube | Zenodo (video + slides)
9.00 am Daniel Smith QC Archive infrastructure updates
Highlights on feature improvements, data artifact generation, and data visualization on the QCArchive website.
YouTube | Zenodo (video + slides)
9.30 am Jeffrey Wagner Updated on core OpenFF infrastructure
New features in the Open Force Field Toolkit, and other developments in core software.
YouTube | Zenodo (video + slides)


Day 2 (Apr 28, 2020)

Time Speaker Talk
8.30 am David Mobley Current Status of OpenFF and our Near-Term Roadmap
I’ll briefly discuss where we are now relative to where we were at the last meeting, and what our near-term horizon looks like for future developments. This will include a brief overview of force field and infrastructure topics of particular interest, with more details to come from other speakers.
YouTube | Zenodo (video + slides)
9.00 am Victoria Lim Benchmark assessment of molecular geometries and energies from small molecule force fields
I will present my results on a comparison of several small molecule force fields in terms of how well they can reproduce geometries and relative conformer energies with respect to QM reference data.
YouTube | Zenodo (video + slides)
9.30 am Jessica Maat QM data selection
QM data selection for OpenFF-1.2.0 force field. Discussion on how the data set was selected to improve the new force field release.
YouTube | Zenodo (video + slides)

Day 3 (Apr 29, 2020)

Time Speaker Talk
9.00 am Simon Boothroyd Which physical properties should we be optimizing against?
An update on refitting the non-bonded (LJ) parameters of the OpenFF force fields against physical property data. A focus is given to exploring which types of physical properties yield the largest improvement in force field performance.
YouTube | Zenodo (video + slides)
9.30 am David Hahn Parsley parameters in protein-ligand binding free energy calculations
QM data selection for OpenFF-1.2.0 force field. Discussion on how the data set was selected to improve the new force field release.
YouTube | Zenodo (video + slides)


Day 4 (Apr 30, 2020)

Time Speaker Talk
8.30 am Michael Gilson How many Lennard-Jones atom types does a force field need?
Results of a data-driven investigation.
YouTube | Zenodo (video + slides)
9.00 am Jeffry Setiadi Host-guest binding calculations with OpenFF Evaluator
Current status of pAPRika integration with OpenFF Evaluator.
YouTube | Zenodo (video + slides)

Day 5 (May 1, 2020)

Time Speaker Talk
8.30 am Owen Madin Future directions in parameterization: Enabling Bayesian inference with surrogate modeling
I’ll discuss our interest in using Bayesian inference to make decisions about non-bonded interactions, and our planned usage of surrogate models to achieve this. I’ll use the 2CLJQ system as a case study.
YouTube | Zenodo (video + slides)
9.00 am Yuanqing Wang Near-Chemical Accuracy at Molecular Mechanics Speed: Neural Parameterization of Extended Force Field using Graph Nets
By using graph neural networks capable of automatically perceiving chemical environment and directly producing parameters for a fast classical potential, we show it is possible to flexibly assign parameters to near-arbitrary biomolecular systems while achieving near-quantum chemical accuracy with molecular mechanics speed.
YouTube | Zenodo (video + slides)
9.30 am Hyesu Jang Update on Parsley minor releases (OpenFF-1.1.0 and 1.2.0)
Valence parameter re-fitting was done with (1) parameter definition adjustments and (2) newly designed QM training set, followed by performance validation.
YouTube | Zenodo (video + slides)


2nd Open Force Field Workshop

We hosted the Second Open Force Field Workshop again in La Jolla, CA on August 30-31, 2019. The first two days of the workshop focused on the progress update and future planning of OpenFF efforts in Year 2. The workshop was extended for another day (Sep 1), which was reserved for OpenFF team discussions, but open to interested Industry Partners. The venue was provided by the Skaggs School of Pharmacy and Pharmaceutical Sciences at UCSD campus.

Read more
Read less

Aug 30-31, 2019 | La Jolla, California

The meeting was open to Consortium members and collaborators, and parts of the meeting were recorded for posting online with accompanying slides. Watch the videos from this workshop on YouTube and access presentation slides on Zenodo.

Please see the full program below:

Agenda

The proposed agenda might be subjected to minor changes before the event.

Day 1 (Aug 30)

Discussion notes for Day 1

Time Event Speaker
7.30 - 8.15 _ Breakfast _
8.15 - 8.30 Welcome OpenFF PIs
8.30 - 9.15 Progress and status for the OpenFF small molecule force field Video
9.15 - 10.00 OFF Toolkit Showcase: Current capabilities Video
10.00 - 10.30 Discussion of progress and current status Video
10.30 - 11.00 _ Coffee break _
11.00 - 11.30 Parameterization perspective I: Parameterization methodology Video
11.30 - 12.00 Parameterization perspective II: Property Calculator Video
12.00 - 12.15 Parameterization perspective III: Data for parameterization and validation - QCArchive Video
12.15 - 12.30 Parameterization perspective IV: Data for parameterization and validation - Physical properties Video
12.30 - 13.30 _ Lunch _
13.30 - 14.30 Open discussion of the morning talks: Feedback and outlook for Y2, near-term planning, challenges Chair: Mike Gilson
14.30 - 15.15 Breakout session I: Force field assessment - physical properties of small molecules Chair: Michael Shirts
14.30 - 15.15 Breakout session I: Chemical perception / addition of new parameters / chemical space coverage Chair: David Mobley
15.15 - 15.30 Breakout session reports
15.30 - 16.00 _ Coffee break _
16.00 - 16.45 Breakout session II: Force field assessment - protein-ligand free energies - [discussion summary] Chair: John Chodera
16.45 - 17.00 Breakout session II: Parameter definition, dataset selection, optimization strategies Chair: Lee-Ping Wang
18.30 **_Workshop dinner:_** ** Rock Bottom Brewery**


Day 2 (Aug 31)

Discussion notes for Day 2

Time Event Speaker
7.45 - 8.30 _ Breakfast _
8.30 - 9.30 Advisory Board meeting (closed)
9.30 - 10.15 Future parameterization perspective: Plan for Year 2 and beyond Video
10.15 - 10.45 _ Coffee break _
10.45 - 11.30 ANI-1 and neural net potentials Video
11.30 - 12.30 Open discussion: Anticipated next force field development steps Video
12.30 - 13.00 Concluding remarks and future outlook
13.00 - 14.00 _ Lunch _
14.00 - 15.15 Free discussion: Short-term and long-term plans for OpenFF Concurrent 1:1 sessions for Partners with J. Wagner
15.15 - 15.45 _ Coffee break _
15.45 - 17.00 Free discussion: ML perspectives in OpenFF Concurrent 1:1 sessions for Partners with J. Wagner


Day 3 (Sep 1)

Internal OpenFF team discussions and planning session.

Time Event Moderator
8.30 - 9.00 _ Breakfast _
9.00 - 10.30 Internal organization and project structure Karmen Condic-Jurkic
10.30 - 11.00 _ Coffee break _
11.00 - 13.00 Roadmap planning and deliverables for Year 2 Karmen Condic-Jurkic
13.00 - 14.00 _ Lunch _
14.00 - 15.00 Summary
15.00 - 17.00 Free discussion and departure

1st Open Force Field Workshop

The First Open Force Field Workshop was held in La Jolla, CA on January 7-8, 2019. The meeting was open to Consortium members and collaborators, and parts of the meeting were recorded for posting online. Watch the videos from this workshop on YouTube and access presentation slides on Zenodo.

Read more
Read less

Jan 7-8, 2019 | La Jolla, California

Please see the full program below:

Agenda

Comments and questions for discussion during the workshop are invited in the Slack channels indicated after each topic; e.g, #project-planning for "Roadmap overview".

Jan 7

Research topics: 45 min talk + 15 min discussion

Time Event Speaker Slack channel
7.45 - 8.00 Introduction [PDF] [Video] Gilson
8.00 - 8.30 Roadmap overview [DOI] [Video] Shirts #project-planning
8.30 - 9.30 Short-term parameter optimization strategy: ForceBalance [DOI] [Video] Wang #forcebalance
9.30 - 10.00 Coffee break
#random
10.00 - 11.00 Long-term parameter optimization strategy: Bayesian Fitting [DOI] [Video] Chodera #bayesian-inference
11.00 - 12.00 Chemical perception and SMIRNOFF typing [DOI] [Video] Mobley/Bannan #smirnoff
12.00 - 1.00 Torsion fitting and fragmentation [DOI] [Video] Chodera/Wang #torsions
1.00 - 2.00 Lunch (catered)
2.00 - 2.30 Group discussion to define ad hoc breakouts and breakout leaders/note-takers

Proposed Breakout Session 1 : typing via chemical perception, torsion fitting

Proposed Breakout Session 2 : useful physical property data, parameter fitting ||

Time Event Speaker Slack channel
2.30 - 3.00 Breakout Session 1
3.00 - 3.30 Coffee break
#random
3.30 - 4.00 Breakout Session 2
4.00 - 5.00 Reconvene for breakout reports and all-hands discussions

Breakout session notes: Benchmarking Accuracy [PDF] |

Breakout session notes: Property calculation framework [PDF] |

Breakout session notes: Use cases [PDF] |

Time Event Speaker Slack channel
5.00 - 5.15 Roadmap recap and discussion (reviewing Slack input from remote attendees)
#project-planning
6.00 Group dinner in La Jolla at Rock Bottom Brewing


Jan 8

Research topics 45 min talk + 15 min discussion

Time Event Speaker Slack channel
8.00 - 9.00 Charge models [DOI] [Video] Schauperl #charge-models
9.00 - 10.00 Fitting non-torsion valence terms [DOI] [Video] Mobley/Wang #valence
10.00 - 10.30 Coffee break
#random
10.30 - 11.30 Software architecture and infrastructure [DOI] [Video] Wagner #infrastructure
11.30 - 12.30 Fitting condensed-phase properties [DOI] [Video] Chodera/Shirts #propertycalculator
12.30 - 1.30 Lunch (catered)
1.30 - 2.00 Group discussion to define ad hoc breakouts and breakout leaders/note-takers

Proposed Breakout Sessions: electrostatics, software preview / toolkit hands-on ||

Time Event Speaker Slack channel
2.00 - 4.00 Breakout Session (with coffee)
4.00 - 5.00 Reconvene and wrap-up discussion

Breakout session notes: neural network potentials [PDF] |