| Bioinformatics Toolbox 1.1.1 Release Notes | |
This section summarizes the new features and enhancements introduced in the Bioinformatics Toolbox 1.1.1.
There are no new features in this release.
The Bioinformatics Toolbox 1.1.1 includes several bug fixes made after the version 1.1 release.
When loading CDF files, the data structure returned by the function affyread has changed. The new structure no longer contains redundant fields. In earlier versions of the toolbox, the X and Y coordinate information for certain probes was not read. This is now corrected and all coordinate information will be read.
To create a consistent format between the functions aminolookup and baselookup, 'letter' was changes to 'code' for the 1-letter abbreviation, and 'code' was changed to 'abbreviation' for the 3-letter abbreviation for aminolookup.
Bioperldemo was originally implemented with Bioperl v1.2.3. Using the current release of Bioperl (v1.4) can cause an error message if you don't have a complete installation of Perl on your system. This happens when MW_BLAST.pl in Bioperldemo is used to generate fresh BLAST results rather than using the sample results included with the Bioinformatics Toolbox. The error message is:
??? Error using ==> perl System error: Can't locate warnings.pm in @INC
In Bioperl v1.4 there is a new function GuessSeqFormat.pm which includes a 'warnings.pm' dependency. This message is simply telling you that Perl can't find 'warnings.pm' in your Perl path. You can correct this error by commenting out the warnings dependency in GuessSeqFormat, or downloading 'warnings.pm' and dropping it in your Bioperl installation directory.
Initializing the matrix now uses log odds ratio probability instead of straight probability.
All hmmprof related functions were updated to accept models with NullX and LoopX transition probabilities.
Null transition probabilities can be loaded from pfam (pfamhmmread) and used to score the model appropriately. A new flag, in hmmprofalign, selects the use of null transitions. By default the NullX probabilities are not considered.
Loop transition probabilities are loaded from the Pfam database, however they are not considered for scoring during alignment. Currently, hmmprofalign does not perform tandem repeat alignments.
A new flag was added to hmmprofalign to activate/deactivate the scoring of the flanks. By default flanks are not scored, however, in order to reproduce HMMER 2.2g results, flanking scores need to be included when this flag is selected.
A third output was added to hmmprofalign to easily recover the domain indexes of the alignment, the output follows the same idea as the output indexed in hmmpromerge. In both cases, Null pointer are now defined by NaN instead of Zero.
| | Known Software Problems | Bioinformatics Toolbox 1.1 Release Notes | |
© 1994-2005 The MathWorks, Inc.