EMAN2

Single particle reconstruction tutorial using demo data. This page is part of the data from the 2011 workshop.

Tutorial

Here are the two parts of the tutorial (linked from the workshop agenda page):

Part 1

Part 2

You MUST use a current version of EMAN2 for these tutorials, not an older release. If you have an old version installed, please upgrade to 2.01 or newer, and be sure to run 'e2bdb.py -c' under the old version BEFORE installing the new version !

EMAN2 will be used throughout the workshop for various purposes and is required to participate in the tutorial sessions. We have just put out the first official release version of EMAN2 (2.0). You can download and install this version from :

http://ncmi.bcm.tmc.edu/ncmi/software/software_details?selected_software=counter_222

installation instructions are provided at the download site. Please make sure you download the correct version for your computer. There are a number of possibilities. If you have any difficulty figuring out which version to download, or installation difficulties, contact gtang@bcm.edu.

Note: We will be doing extensive testing of the tutorials this week. If we discover any last minor bugs, we will release a 2.01 version later in the week, and make an announcement on this page, and on the above software/data list page. You may delay installation until later in the week if this concerns you, but please don't wait until the day before the workshop, or we may not have enough manpower to reply.

Note for advanced users: we strongly suggest using one of our precompiled binaries rather than trying to compile from source, due to the number of dependencies, and limited support resources before the workshop. As all EMAN2 command-line programs are written in python, considerable customization is already possible, even when using the binary distributions. The only users who need to compile from source are those who wish to write new low-level algorithms in C++ or (for the moment) those who wish to experiment with CUDA support.

Required Data Files

The data is provided as either a .zip file or a compressed tar archive. Please download ONE of these files and unpack it into your workshop folder. This file includes all 3 demonstration data sets (GroEL, mm-cpn and Ribosome) :

• eman2_2011_demo_data.tgz

• eman2_2011_demo_data.zip

Once you have installed the software and data, you can test the installation by :

1. Run 'e2speedtest.py'.

   • If it will not run properly, check the installation tips page. Something didn't work right.

   • If it runs properly it will give you a 'speedtest score'. This is a relative indicator of how long it will take to perform refinements on a single core (CPU) on your computer. You can roughly multiply this by the number of cores to get a proper relative estimate. A fast, current generation Intel processor gives a score of ~4400. With 12 cores on a machine (even more will be possible soon), this gives an aggregate score for a single fast workstation of ~50,000. The workstations we will use at the workshop are several years old, and have speedtest scores of ~2000, with 2 cores. A good laptop will give a score of ~3000-3500 and have 2 cores.
2. At the command prompt:
   • cd into the workshop/eman2/test directory and type 'e2display.py'
   • click on each of the 3 files you see in the browser in turn. Each should display an image or volume.

If any of these tests fail, and the installation FAQ doesn't help, please contact sludtke@bcm.edu so he can assist you.

Results data downloads Since the refinements can take some time to run, we have made the complete refinement results for each of the 3 specimens available for download. These results show what you should get if you follow the full reconstruction tutorial to completion. This includes all intermediate files for all steps through the final structure.

These downloads are multiple GB each, particularly the Ribosome. If many people try to download them during the workshop, the network will slow to a real crawl.

• groel-done.tbz (3.7 gb) GroEL refined to ~8.4 A resolution + some demonstrations of resolution exaggeration due to masking and insufficient angular sampling.

• mmcpn-done.tbz (1.4 gb) mm-cpn refined to ~8.4 A resolution

• ribosome-done.tbz (9.5 gb) Ribosome refined to ~11 A resolution with a few different sets of parameters for comparison.

If you don't need to see all of the results in their gory detail, but would like to look through some of the final refinement results, here is a reduced size version for each of these projects. They still have a fair bit of detail without some of the larger intermediate files you are unlikely to look at:

• groel-done-sm.tbz (1.2 gb) GroEL refined to ~8.4 A resolution + some demonstrations of resolution exaggeration due to masking and insufficient angular sampling.

• mmcpn-done-sm.tbz (0.6 gb) mm-cpn refined to ~8.4 A resolution

• ribosome-done-sm.tbz (2.6 gb) Ribosome refined to ~11 A resolution with a few different sets of parameters for comparison.