SATé Settings

This section allows users to control the details of the algorithm. In each iteration, the dataset will be breaking down into non-overlapping sequence subproblems and these subproblems are given to the chosen alignment tool.

  1. There are options under ‘Quick Set’ to allow a more or less intensive search during the SATé iterative process.
  2. ‘Max. Subproblem’ is used to control the largest dataset that are aligned during the iterative process. 
  3. Use the ‘Fraction’ option to express the maximum problem size as a percentage of the total number of taxa in the full dataset.
    This value will be limited by available computational power.
  4. Use the ‘Size’ option for size cutoff in absolute number of sequences.
  5. Select ‘Decomposition’ to choose how the process should be broken to create subproblems.
  6. ‘Apply Stop Rule’ is used to control how SATé should be finished. 
    The decision to stop can be done based on number of iterations (one may be sufficient), the amount of time in hours or ‘Blind Mode Enabled’ meaning that SATé will terminate if it ever completes one iteration without improving the ML score.
  7. Click ‘Start’ to run the SATé analysis.

There will be five files created in the selected directory after SATé is completed.  An alignment file (*.aln), tree file (*.tre), best ML score file (*.score), error file (*.err) and history file (*.out).  Unlike MEGA, SATé does not have bundled tree viewing or alignment viewing programs, so the user will need to open the tree file using one of the tree viewing programs described above and the alignment file using Clustal or a similar alignment viewing program.  The other files can be opened with a text editor (i.e. Notepad, TextEdit). In addition, SATé does not utilize bootstrap testing to support inferred tree topology. Rather, a similarity score from 0-1 (0 is most similar and 1 is least similar) is placed on the branches to aid in topology interpretation.  Hence, other reconstruction methods should be compared to confirm the output.