EukPhylo/PTL2/Scripts-GRID/trimal-trimAl/CHANGELOG

trimAl/readAl v1.3. April 2011.

[General]

  * Rewritten all file format parsers.

  * Added a new file format for compatibility with some programs.
    * phylip_paml

  * Extended some file format parsers.
    * fasta_m10
    * phylip_m10
    * phylip3.2_m10
    * phylip_paml_m10

  * A new control versions system is used: git.

  * Improved general system for tracking builds/revisions.

  * Refactored some portion of codes for improving its legibility and
    comprehension.

[trimAl]

  * Improved HTML functionality. Now trimAl's summary shows which scores have been used
    for trimming the alignment along with scores scales bars.

  * Now it is possible to get an HTML trimAl's summary of previously no-allowed combinations
    of trimming methods. No-allowed combinations were related to the use of methods based on
    several alignments (consistency based scores) and methods based on a single alignment.

  * Some previously no allowed combinations between different trimming methods now are permitted.
    This is oriented to trim a given alignment using scores from several alignments along with
    scores computed from a single alignment.

  * When several alignment are given now it is possible to force the selection of one
    of them using a new option "-forceselect". This is useful to trim a specific alignment
    based on its consistency scores computed from a given set of alignments independently
    if that alignment is the most consistent among the set of those alignments.
    The selected alignment should not be present among alignments in the "-compareset"
    input file.

  * Former "-select" option has been converted into "-selectcols" for removing specific
    columns/range of columns.

  * Added the possibility, using "-selectseqs" option, to remove specific sequences/range
    of sequences from input alignment.

  * Using "-backtrans" option it is possible to trim an input protein alignment and then
    get a back-translated CDS alignment. Coding DNA sequences are mapped before trimming
    to protein sequences through sequences identifiers controlling stop codons, sequences
    lengths, etc, etc.

  * Using "-splitbystopcodon" option is possible, when a back-translation is done, to split
    input coding sequences for its stop codons (at positions multiple of 3).

  * Added a new option "-terminalonly" for removing only columns out of internal boundaries.
    Internal boundaries are defined as the first and last column in the alignment without
    gaps. The columns removal depends on the applied trimming method, that means, if there
    is not columns to be deleted out of internal boundaries, no-columns will be deleted.
    All columns in-between internal boundaries are kept independently of any trimming methods
    decision.

  * Added a new option "-clusters" for getting the most Nth representative sequences from
    input alignment. Representativeness is computed in terms of identity between each pair
    of sequences in the alignment. Then an algorithm is applied for constructing a set of
    clusters based on those identities values. The longest sequences is selected as the
    cluster representative for the final alignment.

  * Added a new option "-maxidentity" for getting sequences representatives at given identity
    threshold. This option allows to get the set of sequences/cluster representatives with an
    identity below to that threshold regarding to rest of the sequences. Sequences with identities
    scores higher than established threshold are clusters to the most similar sequence. The
    longest sequences are selected as cluster representatives for the final alignment.

  * Added a new option "-block" for keeping only columns blocks equal or higher to that size
    in the final alignment. This option is available for manual trimming methods and only for
    "gappyout" method among automated ones.

  * Changed identity scores matrix. Now it is shown "100.000" in the diagonal instead of
    "0". It makes more sense to have a score of 100 for each sequence against itself.

  * Changed options name for making clear their meaning.
    * scc => ssc  - Similarity scores for columns in the alignment.
    * sct => sst  - Cumulative distribution of similarities scores in the alignment.

[readAl]

  * Added new option "-reverse" for getting the reverse of the input alignment.

  * Added new option "-onlyseqs" for getting only sequences from the alignment. This option
    destroys input alignment and returns a fasta file with unaligned sequences.

  * Added new option "-html" for getting an HTML file with alignments residues colored
    according to CLUSTAL scheme. This functionality is similar to SeaView/JalView visualization.

  * Changed option "-format". Now it informs about input file format and if sequences are
    aligned or not.

  * Updated option "-type". Now it informs more about nucleotides datatype: dna or rna.

trimAl/readAl v1.2. April 2009.

  * Added the htmlout to track the trimAl's trimming.

  * Added the possibility to delete spurious sequences from the alignment.

  * Added the complementary and colnumbering options.

  * Added the manual selection of columns.

  * Added one new heuristic method to select the best automated trimming method.
    * Automated1

  * Added two new automated methods to work with gaps and similarity distribution.
    * Strict
    * Strictplus

  * Select only one automated method to work with gaps distribution.
    * Strict => Gappyout
    * Relaxed was deleted

  * Added the version/revisions system.

  * Improved the file format.
    * Added support for phylip 3.2
    * Added support for clustal
    * Added support for fasta
    * Added support for nexus
    * Added support for mega
    * Added support for nbrf/pir

  * Improved of statistics methods.
    * Improved the information of the available methods.
    * Added the identity statistics.

  * Improved of windows size methods.
    * Consistency windows size.

  * Improved the similarity-matrices handling.

  * Improved the support for DNA/RNA alignments.

  * Added readAl.

  * Fixed some bugs from the previous version.

trimAl v1.1. February 2008.

  * Initial public version of trimAl.