EMAN2.3 / SPHIRE 1.2

Release Date

2019-04-24

Summary

This is the 2.3 release of EMAN2. It includes major new additions, including a complete tomography and subnanometer resolution subtomogram averaging workflow(http://eman2.org/e2tomo), greatly improved bispectrum-based classification, upgrading from Qt4 to Qt5 for all of the GUI tools, and a range of other changes and improvements. Major changes are detailed below. Please follow the installation instructions (http://eman2.org/Install/BinaryInstallAnaconda/2.22), particularly for Windows machines, where there is a major choice to make prior to installation.

Description

EMAN2 is a full featured 2-D and 3-D greyscale scientific image processing package with a primary focus on CryoEM and CryoET, but also with uses in other imaging disciplines. It includes a C++ image processing library with hundreds of algorithms for 1-3 D image processing, including registration of extremely noisy data, 3-D reconstruction and a wide range of other tasks. It includes a complete user-level pipeline for single particle analysis and cryoET reconstruction with subtomogram averaging.

Version changes

Major changes in this version include:

• A complete pipeline for CryoET and subnanometer resolution subtomogram averaging. This includes all stages of processing including fully automated fiducial-less tilt-series alignment
• As part of the new subtomogram averaging developments, deep learning based annotation tools have been improved for particle-picking purposes, and a multi-class particle picking GUI is now available
• Significant improvements to deep-learning based tools (based on Tensorflow with GPU support)
• New display filter in e2boxer, which can make particles dramatically easier to see/pick by hand. Very useful when training deep learning picker.
• Focused classification now available in GUI for e2refine_easy. Previously required command-line.
• Improvements/fixes to e2extractsubparticles, an alternative to focused classification
• Further developments in bispectrum-based classification for 2-D and 3-D refinement
• Upgraded user interfaces from Qt4 to Qt5 in preparation for moving to Python 3, hopefully in the next 6-9 months
• Many minor bugfixes