Plugin Hub

Magellon plugins available for install. Browse from your CoreService deployment's plugin manager, or download a .mpn archive directly and upload it through the admin UI.

9 plugins published ctf · 1 fft · 2 motioncor · 1 particle_extraction · 1 particle_picking · 1 square_detection · 1 topaz_particle_picking · 1 two_d_classification · 1

CAN Classifier — 2D classification

v0.1.0

can-classifier

two_d_classification verified

Convolutional auto-encoder-based 2D classification of particle stacks. Reads .mrcs + .star produced by the stack-maker plugin; emits class averages + per-particle assignments + FRC curves.

author: Magellon Team
license: MIT
size: 39.2 KB

Download .mpn JSON

CTFFIND4 — CTF estimation

v0.1.0

ctf

ctf verified

CTF estimation via CTFFIND4. Consumes micrographs from the bus, shells out to the CTFFIND4 binary inside the container, writes the star file under the data plane.

author: Magellon Team
license: MIT
size: 21566.6 KB

Download .mpn JSON

FFT — magnitude spectrum

v1.1.0

fft

fft verified

Computes the magnitude spectrum FFT of an input MRC image and writes the result as a PNG. Pure-Python via numpy; no external binaries required. Pairs with the Magellon CTF and MotionCor plugins as the third "always there" image-processing surface.

author: Magellon Team
license: MIT
size: 289.7 KB

Download .mpn JSON

Magellon Sample Plugin

v0.1.0

magellon-sample-plugin

fft verified

Reference fixture for end-to-end install-pipeline testing. Pure Python with zero runtime dependencies — installs in seconds via uv. Not intended for real cryo-EM workloads; use FFT, CTF, MotionCor archives for that.

author: Magellon Team
license: MIT
size: 4.1 KB

Download .mpn JSON

MotionCor2 — beam-induced motion correction

v0.1.0

motioncor

motioncor verified

MotionCor2-based beam-induced motion correction for cryo-EM movies. Pins ~11 GB host RAM; production deployments need at least a g4dn.2xlarge per the operator sizing notes.

author: Magellon Team
license: MIT
size: 11630.0 KB

Download .mpn JSON

Ptolemy — square + hole detection

v0.1.0

ptolemy

square_detection verified

Ptolemy ONNX models for low-mag square detection + med-mag hole detection. One plugin process serves both categories via separate consumer registrations; manifest declares the square-detection surface as the public face.

author: Magellon Team
license: MIT
size: 5659.2 KB

Download .mpn JSON

Stack Maker — particle extraction

v0.1.0

stack-maker

particle_extraction verified

Extracts boxed particles from a micrograph given picker coordinates; emits RELION-style .mrcs + .star + companion .json. Pure-numpy compute; no external binaries.

author: Magellon Team
license: MIT
size: 20.4 KB

Download .mpn JSON

Template Picker — particle picking

v0.1.0

template-picker

particle_picking verified

FFT-correlation template-matching particle picker. External broker version of the in-process pp/template_picker; deploy this when picker scaling is needed across multiple workers. Two backends now serve PARTICLE_PICKING — operators pick which one to dispatch to.

author: Magellon Team
license: MIT
size: 167.0 KB

Download .mpn JSON

Topaz — particle picking + denoise

v0.1.0

topaz

topaz_particle_picking verified

Topaz CNN-based particle picking (and the companion micrograph denoiser). Two categories share one plugin process — backend_id reflects the picking surface; the denoise category is served by the same image via a second consumer registration.

author: Magellon Team
license: MIT
size: 15331.4 KB

Download .mpn JSON

Programmatic access

CoreService's install pipeline fetches the catalog directly. The same JSON your deployment reads is also public — script away.

/v1/index.json

Full catalog — every published plugin, every version.

/v1/plugins/<plugin_id>.json

Per-plugin detail — same shape as one entry in the index, cheaper to fetch when you only care about one.

CORS is open on both endpoints — fetch directly from a browser without a proxy.