Principle of the FTHOM algorithm (H. Hegyi and S. Pongor, Comput. Applic. Biosci.,9, 371-372, 1992.)

This scoring is repeated for each alignment in the BLAST output and a sumary list is presented in terms of domain homologies, ranked by score. The below example is the analysis of the human C1S light chain, a serin protease which has a propetide removed by proteolysis.

                                Feature                 FT    Score
                                 name                  freq
      
                             DOMAIN SERINE PROTEASE.    28    6697
                                   DOMAIN CATALYTIC.     1     301
                               SIMILAR WITH TRYPSIN.     1     269
                                  DOMAIN BETA CHAIN.     1     256
                             DOMAIN SERINE-PROTEASE.     1     122
                                   DOMAIN SEGMENT A.     1     112
                                  DOMAIN SEGMENT B2.     1      72
                                  DOMAIN SEGMENT B1.     1      62
                                   DOMAIN SEGMENT C.     1      40
                                PROPEP PROLINE-RICH.     3      27
       SITE MAY BE REMOVED BUT IS NOT NECESSARY FOR      1      19
                                         ACTIVATION.
                                            CA_BIND.     1       8
                  SITE CLEAVAGE SITE (BY FACTOR XA).     3       3
      SITE CLEAVAGE (BY FACTOR XIA) (BY SIMILARITY).     2       2
                      SITE CLEAVAGE (BY FACTOR XIA).     2       2
                                  BINDING SUBSTRATE.     2       2
                       SITE CLEAVAGE (BY FACTOR XA).     1       1
                          PROPEP ACTIVATION PEPTIDE.     1       1
            SITE CLEAVAGE BY FACTOR XA, FACTOR XIIA,     1       1
                      SITE THROMBIN ACTIVATION SITE.     1       1
      

The colors indicate that the different domain names were used in the early version of Swiss-Prot used for the search. Since 1992, standardization of domain names started within the SBASE project, and the standard names greatly simplify the output:

    With SBASE 3.0 standard names
    
                                SERINE-PROTEASE.    36 7958
                                PROCESSING-SITE.    14   40
                            PROPEP PROLINE-RICH.     3   27
                                     CA-BINDING.     1    8
    

Note that even a short feature, like a proteolitic procesing site of 1-2 amino acids can be noticed with this simple procedure. FTHOM is part of the domain server (domain@hubi.abc.hu)