Parameter Description

  • PSIBLAST / BLAST: Depending on what you set, that program will be used for searching the sequence database. In case PSIBLAST does not converge, BLAST will automatically be executed.

  • DIP ON: The DATABASE of INTERACTING PROTEINS is one of the databases queried. Sometimes however, you may want to exclude it from your search because of the small size of the database.

  • EVIDENCELEVEL 0: Currently there are nine steps used to populate the master table during the function calculation. However, in many cases all these nine steps may not return results. In that case the functions that came from the maximum number of steps are used for Bayesian Scoring. It may happen that the function(s) that came up in the list are inherently present in larger numbers in the database and therefore need to be viewed with caution. In such a scenario, you could degrade the evidence level to 1 and include the next-best functions that came up in the hit. You can find the list of functions and their overall frequencies here. These frequencies are over all the databases used in this program.

  • MERGELEVEL 0: This parameter works in conjunction with the FULLONTOLOGY parameter. When the FULLONTOLOGY parameter is set "YES" you can use the MERGELEVEL. This basically truncates the full ontology of a function term to the desired depth. If you say MERGELEVEL 5, then it will truncate the depth to 5 and all full ontologies having common terms beyond depth 5 will be merged to depth 5. This helps in merging two functions which are basically different in their teminal term, but same at some other level. Example : Function lineage A: 1:2:3:4:5; Function lineage B: 1:2:3:4:6. On truncation (or MERGELEVEL 4) both A and B are 1:2:3:4 and they are the same function lineage.
    However, full ontology computations are computationally expensive and therefore turned off on the webserver. 1,2,3,4 ... refer to function terms. In reality 1 is always GO:0003674 when dealing with molecular function. The merging operation allows one to get functions at a less specific description, but with higher confidence.

  • FULLONTOLOGY NO: Only the terminal value (basically, the annotation as given in the database) of the functional lineage is used. A full ontology of a molecular function consists of all the paths to the parent term "Molecular Function: GO:0003674".