FSST release notes
==================

Changes in FSST 3.1
- updated bundled data files to UniProt release 15.3 and Gene Ontology 2009.04

Changes in FSST 2.1
- updated the bundled data files to UniProt release 13.3 and Gene Ontology 
  2008.04
- show progress information while building the database

Changes in FSST 2.0
- updated the bundled data files to UniProt release 12.8 and Gene Ontology 
  2008.01
- added progress monitor;
  Current work progress is written to the console in configurable intervals, 
  information provided: percentage complete, estimated time left, and elapsed 
  time.
- added support for external database servers, e.g. MySQL;
  Instead of the embedded database, FSST can now use external database systems 
  that provide an ODBC compliant driver. The driver jar file has to be added
  to the classpath when building the database or running FSST. Use --help to
  get information on the required command line options.
- added support for CCscore, funSimAll, and rfunSimAll;
  CCscore is equivalent to BPscore and MFscore for cellular component. funSimAll
  and rfunSimAll take all three ontologies into account.
- changed output formats to include new scores;
  Abbreviation for "average" has been changed from "ave" to "avg"
  The new column order for a pairwise comparison is:
  acc1, acc2, rfunsim (max), funsim (max), rfunsimAll (max), funsimAll (max), 
  mfscore (max), bpscore (max), ccscore (max), rfunsim (avg), funsim (avg), 
  rfunsimAll (avg), funsimAll (avg), mfscore (avg), bpscore (avg), 
  ccscore (avg), matching pfams, non-matching pfams
  The new column order for an all-against-all comparison is:
  acc, rank, acc, rfunsim (max), funsim (max), rfunsimAll (max), 
  funsimAll (max), mfscore (max), bpscore (max), ccscore (max), matching pfams, 
  non-matching pfams, rank, acc, rfunsim (avg), funsim (avg), rfunsimAll (avg), 
  funsimAll (avg), mfscore (avg), bpscore (avg), ccscore (avg), matching pfams, 
  non-matching pfams
- new matrix conversion tool;
  de.mpg.mpi_inf.bioinf.gotaxexplorer.cml.ConvertMatrix performs 
  post-processing of FSST output files in matrix format. 
  ConvertMatrix allows to:
  - split the matrix and save the column ids, row ids, and the values in 
    different files
  - bring the matrix into symmetrical form by sorting the rows and columns, 
  	saved in new file
  - convert similarity values to distance values and vice versa
  - replace NaN values with a given string, e.g. 0.0 or 1.0