Good practice

When analyzing tomographic data using our in-house software many files of many types are generated that are not intrinsically linked to each other. For a lack of a record management system it is therefore essential that one keeps their analysis organized, such that it can be easily understood and repeated once time has passed. The following instructions are suggestions and examples of how this has been handled in the past.

Table of Contents Reconstructions Naming Subtomogram alignment Folder Structure

Reconstructions

Naming

When naming reconstructions and reconstruction config-files, it is suggested to include resolution, oversampling, supersampling and ctf-mode in the filename to ease later processing. For example as:

 tomo_2_2k_oss5_ss1_ctf1.cfg
 tomo_2_2k_oss5_ss1_ctf1.em

Reconstructions and config-files should always have the same name for easy identification. Once a config file has been created using the reconstruction assistant of Clicker, config files should only be modified in batch when working with large amounts of data. For this purpose, the matlab function modify_cfgs is useful:

artia.cfg.modify({'oldcfg.cfg'}, {'newcfg.cfg'}, ...
               'Parameter1', {'Value1'}, ...
               'Parameter2', {'Value2'});

Subtomogram alignment

Folder Structure

When performing subtomogram alignment using SubTomogramAverageMPI the folder structures and file name conventions below have been established by members of the group. Using this strategy has a number of advantages:

• The OtherInputs-folder always contains all necessary volumes within the averaging project
• The config files that correspond to a certain iteration canbe easily identified
• Specifying box size, radius of the sphere, gaussian sigma, wedge angle, etc. in the filename makes identification, re-use and re-creation of mask/maskcc/wedge-files easier/possible.
• By leaving the parts folder a level above the motls/refs-folders disk-space can be saved by avoiding duplication, while still maintaining order and separating different alignment strategies.

 n - Number of iterations             g  - Sigma of gaussian
 
 j - Tomogram number                  a  - Missing angle
 
 i - Particle number                  s# - start iteration
 
 b - Box size                         e# - end iteration
 
 r - Radius
 
 '''Foldername'''                           ''Filename''