Getting Started#

cryosparc-tools is an open-source Python library that enables scripting access to the CryoSPARC cryo-EM software package.

CryoSPARC Architecture with cryosparc-tools

Use it for the following use cases:

  • Programmatically read and write exposure, particle and volume data

  • Access project, workspace and job data

  • Build and run jobs to orchestrate custom cryo-EM workflows

  • Extend CryoSPARC functionality with third-party software packages


This guide documents usage of the cryosparc-tools Python library. For CryoSPARC installation or general CryoSPARC usage instructions, read the official guide.

Source code is available on GitHub.


CryoSPARC installation must be accessible via one of the following methods:

  • Running on the local machine

  • Running on a machine on the same network with BASE_PORT + 2, BASE_PORT + 3 and BASE_PORT + 5 open for TCP connections

  • Running on a remote machine with BASE_PORT + 2, BASE_PORT + 3 and BASE_PORT + 5 forwarded to the local machine

See SSH Port Forwarding documentation for accessing a CryoSPARC instance on a remote machine accessible via SSH.

cryosparc-tools is only available for CryoSPARC v4.1 or newer. If using CryoSPARC v4.0 or older, please see the Manipulating .cs Files Created By CryoSPARC guide.


Install cryosparc-tools in the command line from PyPI:

pip install cryosparc-tools


Use the version of cryosparc-tools that corresponds to your CryoSPARC minor release version. i.e., if the CryoSPARC version is vX.Y.Z, use the latest vX.Y tools package. The Z component does not need to match.

For example, if you are running CryoSPARC v4.1.2, install cryosparc-tools with pip install cryosparc-tools~=4.1.0 (equivalent to pip install "cryosparc-tools>=4.1.0,<4.2"). If you later update to CryoSPARC v4.2.0 or v5.0.0, re-install the corresponding tools package with pip install cryosparc-tools~=4.2.0 or pip install cryosparc-tools~=5.0.0 respectively.


Import from a Python module and connect to a CryoSPARC instance. Include your CryoSPARC license ID, the network hostname of the machine hosting your CryoSPARC instance, the instance’s base port number and your email/password login credentials.

from import CryoSPARC

cs = CryoSPARC(

This assumes CryoSPARC base ports +2, +3 and +5 (e.g., 39002, 39003 and 39005) are available at localhost on the local machine. If CryoSPARC is on another machine on the same network with a different host, say “hostname”, use host="hostname".

Query projects, jobs and result datasets:

project = cs.find_project("P3")
job = project.find_job("J42")
micrographs = job.load_output("exposures")

for mic in micrographs.rows():

Load and save datasets directly (assumes project directory is available on current machine):

from cryosparc.dataset import Dataset

path = project.dir() / "J43" / "particles.cs"
particles = Dataset.load(path)

shift_y, shift_x = particles["alignments2D/shift"].T
mic_shape_y, mic_shape_x = particles["location/micrograph_shape"].T
new_loc_x = particles["location/center_x_frac"] * mic_shape_x - shift_x
new_loc_y = particles["location/center_y_frac"] * mic_shape_y - shift_y
particles["location/center_x_frac"] *= new_loc_x / mic_shape_x
particles["location/center_y_frac"] *= new_loc_y / mic_shape_y

Running the Examples#

The example Jupyter notebooks require additional dependencies to run. Use conda to create a new Python environment with the required dependencies. Here the environment is named cryosparc-tools-example but any name may be substituted:

conda create -n cryosparc-tools-example -c conda-forge python=3 numpy==1.18.5
conda activate cryosparc-tools-example
pip install matplotlib~=3.4.0 pandas==1.1.4 cryosparc-tools

For speed, these do not include the dependencies for the crYOLO example notebook. Optionally install crYOLO with these commands:

conda install -c conda-forge pyqt=5 libtiff wxPython=4.1.1 adwaita-icon-theme
pip install nvidia-pyindex
pip install "cryolo[c11]"

Install Jupyter with this command:

pip install notebook

Example notebooks ran on Ubuntu Linux with x86-64 bit architecture.

Next Steps#

Browse the included examples real-life use cases for cryosparc-tools. Read the API Reference for full usage capabilities.


For questions, bug reports, suggestions or source code contributions, please read the contribution guide.

If you publish an open-source tool that uses this package to GitHub, add the cryosparc-tools topic to your repository so others may discover it. Browse tagged packages here.


cryosparc-tools is licensed under the BSD-3-Clause license. View full license text.