d-SEAMS

Author:

d-SEAMS core team

Read

Load a LAMMPS frame and print nop, frame, and box.

Quickstart
Classify

CHILL+ four-neighbour star: cubic, hexagonal, interfacial, clathrate, interClathrate, or water.

Classify ice on the mixed TIP4P example
Score

Cage membership: HC is ice Ih, DDC is ice Ic, else water.

CHILL+ versus cages

Overview

d-SEAMS 2.14 reads a molecular-dynamics trajectory and names the ice in it. The name is Deferred Structural Elucidation Analysis for Molecular Simulations. Release 2.14.0 of this repository is the C++ engine libyodaLib, the C API libseams (include/seams.h), the header-only C++ API include/seams.hpp (it links libseams), the C ABI libseams_lang (bindings/cxx/seams_lang.h, used by the Python, R, and Julia fronts), the seams command, and the LAMMPS, GROMACS, OpenMM, PLUMED, and in-situ drivers. gmx-seams reads the trajectory gmx mdrun wrote. The standalone PLUMED action (dseams-plumed v0.4.1) is indexed in PLUMED-NEST egg 26.018. seams --version prints 2.14.0.

See the PLUMED page for the plugin.

CHILL+ labels an oxygen from its four-neighbour star: cubic, hexagonal, interfacial, clathrate, interClathrate, or water. Cage membership is a separate score. A molecule in a hexagonal cage (HC) is ice Ih. A molecule in a double-diamond cage (DDC) is ice Ic. A molecule in neither is water. Primitive rings and Steinhardt \(q_l\) are computed in the same engine. CHILL+ versus cages records where the two scores disagree.

A source build from this tree, when the compilers are already on PATH, is meson setup build && meson compile -C build. pixi is pixi run setup && pixi run build. Install has each method.

The ice Ih frame shipped in this tree is input/traj/ice1h.lammpstrj, 128 molecules of LAMMPS type 1.

seams read input/traj/ice1h.lammpstrj
seams chill-plus input/traj/ice1h.lammpstrj --cutoff 3.5

seams read prints nop 128 frame 1 box 15.6457 14.7071 18.073. seams chill-plus prints nop 128 hexagonal 128. The quick start classifies an eight-atom cubic cell from C and from Python. From a trajectory to the ice front adds the ring census on this frame and the ice front on input/traj/interface_f0.lammpstrj.

Each other language is its own repository. Use a language binding names the package and the library each one links.

  • Python: pydseamslib (import dseams, import pydseams), with pip install 'pydseamslib==2.14.0' from PydSEAMSlib

  • Lua and Fennel: require("dseams"), clone yodaStruct and run luarocks make in the checkout

  • R: remotes::install_github("d-SEAMS/dseams-r") from dseams-r

  • Julia: Pkg.add(url="https://github.com/d-SEAMS/Dseams.jl") from Dseams.jl

  • Rust: cargo add dseams@2.14.0 from dseams-rs

  • Fortran: fpm Git dependency dseams from dseams-fortran

The 2.14.0 notes are the changes since 2.13.0. Migrate from 1.x maps yodaStruct -c onto seams.

Suite stack

        flowchart TD
  DUMP["LAMMPS dump and XYZ"] --> LIB[libyodaLib]
  NB["linkcell or vesin"] --> LIB
  LIB --> CLI[seams]
  LIB --> API[libseams]
  LIB --> ABI["libseams_lang<br/>Python, R, Julia"]
    

libyodaLib reads a LAMMPS dump or an XYZ file. chemfiles (.pdb, .gro, .dcd) and the .con reader are compiled in when those libraries are found. The cutoff neighbour list is linkcell when that library is built, and vesin otherwise. seams --features prints which backends this binary has. The seams command, libseams, and libseams_lang each link libyodaLib. include/seams.hpp links libseams. Python and Julia link both libraries. R links libseams_lang. Lua, Rust, and Fortran link libseams.

Documentation structure

The user guide groups the tutorials and the how-tos by task. The explanation pages are the background for those tasks.

Getting started

Install the engine and classify one shipped frame.

Getting started
User guide

Tutorials and how-tos, grouped by the analysis you want to run.

User guide
API reference

The C API, the C++ API, the command line, and the language bindings.

API reference
Project

Release notes and how to cite d-SEAMS.

Changelog

The C++ header reference is *api-cpp*.

Performance

The numbers for OVITO, freud, ClavisCage TRACE, SQQ, and PLUMED are on Head-to-head. The sitting medians are on the sitting tables.

License

MIT. Cite the 2020 JCIM paper (DOI 10.1021/acs.jcim.0c00031). See Citation.