atomview 0.1.1

Lightweight molecular structure viewer for Jupyter notebooks using py3Dmol

atomview

A lightweight, standalone molecular structure viewer for Jupyter notebooks. Visualize biotite AtomArray structures interactively using py3Dmol (3Dmol.js) — with chain-coloured cartoons, ligand sticks, ion spheres, surfaces, and hover labels out of the box.

atomview extracts and refactors the excellent view() function from atomworks (atomworks.io.utils.visualize) into a focused, minimal package. If all you need is a quick way to render structures in a notebook, you no longer have to pull in atomworks and its full dependency tree (RDKit, ML tooling, dataset infrastructure, etc.).

Installation

pip install atomview

Or with uv:

uv add atomview

Quickstart

# from urllib.request import urlretrieve
# urlretrieve("https://files.rcsb.org/download/4HHB.cif", "4hhb.cif")

from atomview import load_structure, view

structure = load_structure("4hhb.cif")

view(structure)

Gallery

Default protein view	Surface overlay
One-line rendering with automatic chain colouring.	Optional translucent VDW surface for shape context.

Ligand focus	Mixed complex
Zoom to a ligand or binding-site selection.	Proteins, nucleic acids, ligands, and ions styled together.

API

view()

The main entry point. Returns a py3Dmol.view object that renders inline in Jupyter.

from atomview import view

v = view(
    structure,
    show_cartoon=True,            # cartoon for polymers
    show_surface=False,           # VDW surface overlay
    show_hover=True,              # hover labels
    hide_solvent=True,            # remove water
    hide_crystallization_aids=True,
    zoom_to_selection=None,       # e.g. {"chain": "A", "resi": 35}
    width=600,
    height=400,
    colors=None,                  # custom colour list, or use defaults
)

Why not just use atomworks?

atomworks is a fantastic, comprehensive framework for biomolecular modelling maintained by RosettaCommons. But it's a big library — it bundles I/O, transforms, ML dataset tooling, RDKit integration, and more. If you just want to call view(structure) in a notebook, that's a lot of overhead.

atomview gives you exactly that one thing: a great molecular viewer with sensible defaults, in a package that installs in seconds.

License

MIT