BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260421T090513Z
LOCATION:Plenary Room (Bldg. 6 - 001)
DTSTART;TZID=Europe/Stockholm:20260629T193400
DTEND;TZID=Europe/Stockholm:20260629T193500
UID:submissions.pasc-conference.org_PASC26_sess124_pos113@linklings.com
SUMMARY:A Flexible Interface for Neural Network Potentials in GROMACS
DESCRIPTION:Lukas Müllender and Berk Hess (KTH Royal Institute of Technolo
 gy) and Erik Lindahl (KTH Royal Institute of Technology, Stockholm Univers
 ity)\n\nWe present a new interface for hybrid machine learning/molecular m
 echanics (ML/MM) simulations implemented in the molecular dynamics engine 
 GROMACS. The interface enables neural network potentials (NNPs) trained in
  the PyTorch framework to contribute energies and forces during molecular 
 dynamics (MD) simulations, either for selected subsystems or entire molecu
 lar systems. By defining a flexible set of model inputs and outputs, the i
 nterface is agnostic to specific NNP architectures and can accommodate a w
 ide range of descriptor-based and message-passing models.\nThe design inte
 grates NNP inference into established GROMACS workflows while remaining co
 mpatible with advanced sampling and free energy methodologies. We demonstr
 ate the capabilities of the interface using several representative applica
 tions, including solvation structure calculations, enhanced sampling of pe
 ptide torsional free energies, absolute solvation free energy calculations
 , and protein-ligand binding simulations. Across these examples, ML/MM sim
 ulations reproduce established reference results and, in some cases, impro
 ve upon classical force field descriptions, at substantially reduced cost 
 compared to QM/MM approaches.\nThe interface is available in recent GROMAC
 S releases and provides a practical foundation for incorporating machine l
 earning potentials into production MD simulations, with ongoing developmen
 t aimed at extending embedding schemes and improving performance and scala
 bility.\n\n
END:VEVENT
END:VCALENDAR