v1.0.6

This is a minor update to reflect recent changes.

Bug fixes

• #14 - mmseqs parameter are properly adjusted for nucleotide searches - df9d3e6
• #21 - Genes with 3 or less uniquely aligned sequences will now be excluded from phylogenetic inference - 16a45c1

Adjustments

• Unconventional symbols now can pass through mafft aligner - 581f881
• #28 - Pipeline will track down the real path if symbolic links are given - f426ac3

v1.0.5

New features

• Build from source is now available - fc3578b
  • Any builds including commits between stable releases will be marked as unstable, and will be warned - 63240fd
  • Powered by appassembler by mojohaus
• profile - Validation sensitivity adjustment feature - #19, 6dbbe3d
  • --sensitivity 3 (default): Use original algorithm (cov. 80%, 50%, e-value only)
  • --sensitivity 2 : 2-step search (cov. 80%, 50%)
  • --sensitivity 1 : 1-step search (cov. 80%)
• New mascot! Special thanks to Gallois - Image source, 3e2e9a9

Adjustments

• profile, profile-rna, profile-pro - Pipeline will automatically download config files for its initial run - #16, 75265d5
• profile - Use default MMseqs easy-search parameters for --search-type and --cov-mode flags - 3a1fabf
• download - Use the current working directory for target sample - 64ac3ee
• align - Dependency can now be internally set through arguments - 03f11ae

Bug fixes

• profile, profile-pro - Completely synchronized output for multi-threading - 1f8a7d7
• Minor fixes including typos

v1.0.4

New features

• Provided better usability to tree and align - 557de11, e384942, 2d9fbfd
  • The modules will no longer produce a sub-directory with random hexadecimal string names even when the run name is not given.
  • Instead, outputs will be directly written in the given output directory.
  • tree module will no longer require -l option. label will be used as a default parameter.
• Introduced file logging based checkpoint feature - #12, a1e9177
  • Currently applied to tree and align

Bug fixes

• Fixed mmseqs detection in interactive profile mode - #9, 35e52bb
• Faster and more reliable reverse complement generator - #13, fe699d1
  • Thanks to @Changwanseo
• Bumped Maven JSON dependency - 7de7902
• Minor fixes

v1.0.3

Notifications

From this version, ufcg will be published as a bioconda package.

New features

• ufcg binary - 08f9efd
  • Redirected java -jar command into a simple shell script binary for further convenience
• convert module - b53e5e7
  • Converts .ucg file into a FASTA file by compiling its nucleotide/protein sequences with designated headers
• download module - a1d7504
  • Allows for downloading search databases to drastically reduce the package size
• align and tree modules now only includes single-copy genes for downstream analyses - #4, 6424f44
  • -c option introduced for these modules, which allows for multiple copies to be analyzed
  • Thanks to @BrigidaGallone

Bug fixes

• Removed file deletion script using java.lang.ProcessBuilder, replaced with java.io.File - #2, 6c6656b
• Replaced substring matching to entire string matching in FASTA parser - #3, 19f9826
• Timeout wrapped around augustus run to prevent excessive resource consumption - #5, d6a4742
• Stabilized profile interaction module - 01dd389
• Minor fixes on program options and prompt messages

Notes

Special thanks to @Sebastien-Raguideau for their helpful reviews and suggestions!

v1.0.2

• Smarter and permissive batch input detection
  • Allows any discrepancies between the actual file system and the information given by the metadata
• Sanity check routine for augustus/fastBlockSearch binaries
• Fixed issue where file reader stalls at the pre-handled fasta header
• Null-safe code blocks for temporary file mapper
• Code cleanup
• Include publication
• Minor fixes

v1.0.1

• augustus v3.5.0 dependency and compatibility
• Fixed behavior of profile batch mode that skips to drag the contig if its name is overlapping with those from previous files
• Minor fixes

v1.0

Initial release