Scipion integrates existing software allowing you to smoothly switch from one package to another offering endless combinations. We started with single particle analysis (SPA) methods but we are currently expanding our toolset to include atomic modelling and tomography. The following graph is dynamic and clickable, you can play with it!See all plugins
One of Scipion’s main goals is to provide an entire track of the processing pipeline. Each new method execution is stored, together with selected parameters. All output logs are also stored, allowing the expert user to inspect more deeply the underlying commands and their results. You will always know what have you done!.
Users interact with Scipion through a collection of GUIs that provides a uniform interface for a plethora of heterogeneous EM programs. Also, some parameters have “wizards” - special interfaces that allow the adjustment of parameter values while showing their effects in real time.
Scipion is now a plugin framework. This will speed up the release of new packages or new updates through a plugin manager, without the need of releasing a new Scipion version. Plugins can also install the actual image processing software "behind", if the license allows it.
The coexistence of several software packages in the same workflow engine allows the comparison of different estimates of the parameters used for the reconstruction, e.g., we can compute the consensus at the level of CTF, identified particles, particle classification, angular assignment, etc. From the consensus it is easier to identify biases in the parameter estimates and reduce their variance.
Scipion is very well integrated with High-Performance Computing environments (fat nodes, clusters, cloud, ...). Jobs can be submitted to queues, monitored and stopped. There are also configuration modes in which the amount of required GPUs for the preprocessing steps can be minimized.
Scipion is very rich in the amount of tools available to verify the quality of the reconstructed volume (comparing to experimental SAXS curves, to previous PDBs, comparing reprojections, comparing different reconstructions among themselves, evaluating the alignability of the particles used and the amount of overfitting, ...)