SAUZEROR

Structure Alignments Using Z-scaled EigenRanks Of Residues

Implementation of the novel EigenRank approach by Heinke et al..

python sauzeror.py

Example

Align several domains, listed in domains.txt, with every structure in the SCOPe directory and save the output under results.txt:

python sauzeror.py -v align domains.txt ../SCOPe/ -o results.txt

Options

Run sauzeror.py without arguments to view a help message.
align mode aligns two sets of protein structures.
profile mode saves EigenRank and LeaderRank profiles along with atom coordinates and residues to a .profile file.

Tips

• Please only use this tool in a sort of batch situation. Importing and firing up multiprocessing can take about a second. numba compiles the machine code for another half second the first time the script is runs.
  No worries though: if you have some hundred or thousand structures, each alignment will take only milliseconds.

• Use atomium if you have anything but normal .pdb files (i.e. mmcif) or if you only have the PDB-IDs for it to fetch. The primary parser works fine otherwise and is faster.

• Only the first chain of the first model is selected of each structure file. There are two options should you want to analyse more multiple chains per structure:

    1. split the file so that there's only one chain per file
    2. change the script to use all chains
  

For assistance with SAUZEROR email me at [email protected]. I'm also open to any discussion regarding the script or it's underlying method. Variants of SAUZEROR and other stuff is at https://github.com/juvilius/sauzeror-etc.

Requirements

• Python
• numpy
• scipy
• numba
• matplotlib (for graphical output)
• atomium (as a parser option)