From Dynamo
Revision as of 13:28, 30 August 2023 by Daniel Castaño (talk | contribs) (→‎Beta Version)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Dynamo package

Matlab and Linux Standalone

Please follow this google drive link to download Dynamo. You can safely ignore any complaint of your browser about not being able to preview the file.

This package is around 8Gb as it includes the standalone Matlab MCR libraries of Matlab R2021a. The tar file can be installed both as standalone or inside the Matlab environment.

Standalone for MAC

If you plan to use Dynamo in the Mac without a Matlab license use this google drive link. This link is updated less frequently than the link for the latest Linux version. If you want the latest version of Dynamo as standalone for the Mac, drop us an email.

Standalone for Windows

A standalone/matlab version for Windows 7 can be downloaded in this google drive link. This link is updated less frequently than the link for the latest Linux or Mac versions. The Windows version is unsupported and distributed "as is". If you find installation problems, we advice you to use a different platform.

Beta Version

This Google Drive link contains our newest version, 1.1.539 without MRC libraries. It will work under Matlab.

Example data sets

Examples of training data sets will be updated here

Direct link

You can download the link from the command line (no need to access google drive via browser) in Linux with this command:

wget --load-cookies /tmp/cookies.txt "$(wget --quiet --save-cookies /tmp/cookies.txt --keep-session-cookies --no-check-certificate '$DYNAMO_LINK' -O- | sed -rn 's/.*confirm=([0-9A-Za-z_]+).*/\1\n/p')&id=$DYNAMO_ID" -O dynamo-v-1.1.509_MCR-9.6.0_GLNXA64_withMCR.tar && rm -rf /tmp/cookies.txt

note that you will need to replace 1BpZ9pv3Zq5_jSOrIqotoqijOKpWMUQ4a with the actual value in the google link, likewise for the name of the tarball (thanks for Samuel Fux at ETH Zürich for this tip).

Version features

We are restructuring our modules for alignment of tilt series and classification. Next stable releases will come in winter 2023; contact us if you want to be a beta tester.


  • IMOD-compliant tilt angle convention.
  • CTF hand check in automated tilt series alignment.
  • GPU and multiGPU enabled template matching.


  • Bug with opening of Matlab parallel pool corrected for standalone and Matlab.
  • Enhanced tools for manual alignment of tilt series through fiducial markers.


  • Correction of problems derived from the different behaviour of dread and dwrite towards extension .txt.

In previous versions, a file with extension .txt would be read as a matrix of number. From version 1.1.516, this default was changed into a cell array of text lines (to make the extension coherent with the usual interpretation of such files). This has cascaded some backward compatibility problems. This version solves all the reported ones; please let us know if you identify further bugs.


  • Debugging of problems with standalone.
  • Phase randomization tools.


  • Debugging of PCA system.
    • PCA works now on GPU, and includes a new graphics system.
  • New version of </t>dmapview


  • Debugging of problems in the mesh system.


  • tilt series alignment: improved reload of workflows, allows selection of reconstruction height
  • PCA: new internal algorithm. Reworked workflow for command line


Corrects a bug introduced into 1.1.454 concerning the multicore execution.


Includes the Particle List File object (to access particles that are not organized as classical data folders)


Tilt series alignment.