
Prokka ChangeLog
----------------

v1.12 - 14 Mar 2017
* Lots of bug fixes.
* Improved documentation, especially installation
* Newer static binaries for Linux and MacOS
* Updated protein and ncRNA databases
* Better default unique locus_tag based on contig MD5
* Option --cdsrnaolap to allow tRNA/rRNA overlapping CDS
* Option --addmrna to add mRNA features for each CDS
* Option --noanno to just predict CDS coordinates but not /product
* Option --fast to just BLAST sprot and ignore slower HMMs
* Better log file
* New .tsv output of annotated features

v1.11 - 15 Feb 2015
* Option --proteins now supports .GBK/.EMBL files directly!
* Support for user supplied HMM via --hmms [Connor Driscoll]
* Add Prodigal -c option when in metagenome mode [Daan Speth]
* Wierd coordinate errors with long Spades contig names [Stephan Pabinger]
* Replaced dodgy OSX aragorn binary [Yevgeny Nikolaichik]
* Handle translation table 25 [Mads Albertsen]
* Fix semantically incorrect GFF3 output [Marc Hoeppner]
* Fix contigs with pipe characters from Quiver [Vaughn Cooper]
* Keep original contig ordering [Jane Hawkey]
* Fix protein translation with alternate --gcode [Andreas Leimbach]
* Various bug fixes in prokka-* support scripts [Andreas Leimbach]
* Workaround tbl2asn bug with COORDINATES:profile [Wiep Smits]
* Provide MD5 checksum for website download [Willem VA Viljoen]
* Hopefully slightly improved cleanup_product() function
* Ensure released are tagged on github for Linux Brew [Shaun Jackman]
* Check databases exist before searching against them
* Updated bundled binaries

v1.10 - 28 July 2014
* Support for barrnap 0.4 with Archaea and Mito support [Lionel Guy]
* New legal assembly_gap feature in gbk, sqn, tbl files [Connor Skennerton]
* Bug fix: for hypos with /gene and /EC [Connor Skennerton]
* Improved cleanup_product [Connor Skennerton]
* Bug fix: aragorn tmRNA parsing [Connor Skennerton]
* New --depends option to list all dependencies
* Check for exes needed for --setupdb before attemping it [nsoranzo]
* Bug fix: missing ID on 'gene' feature tytpes [Chris Fields]
* Upgraded minced to 0.1.6 to avoid out-by-1 bug
* Bug fix: Bio::RootI file handle issues in latest BioPerl [Chris Fields]
* Bug fix: --setupdb doesn't have bundled binaries in PATH [Connor Skennerton]
* Command line and other improvements [Shaun Jackman]
* Added --rnammer switch to prefer RNAmmer over Barrnap 
* Added --rawproduct to turn off /product cleaning
* Typo in help for --compliant [Haruo Suzuki]

v1.9 - 8 May 2014
* MAJOR: Bug in prokka-uniprot_to_fasta_db: was allowing Altname instead of Recname
         This affected the /product names that were used in the core protein database!
* Updated kingdom databases (sprot)
* Updated 'cmscan' binaries from 1.1rc to 1.1 (linux + darwin)
* Fixed bug when input fasta has duplicate sequence IDs [James Doonan]
* Removed linux/nhmmer binary as it is not used by Prokka
* Experimental --kingdom Mitochondria
* prokka-uniprot_to_fasta_db now has --term empty as default
* More prokka-XXXX_to_fasta_db scripts (not for general use just yet)
* New --setupdb function to press/index databases ready for Prokka to use
* New --cleandb function to remove all redundant indices (useful for making tarball)
* Introduction of 'HAMAP' HMM before CLUSTERS as smaller, faster, reliable
* Removal of 'Cdd' HMM as no longer maintained
* Removal of 'TIGRFAMs' HMM as doesn't add much
* New dependencies of hmmpress and cmpress to support --setupdb
* Fixed bug with parsing "X.Yb" version format with signalp [Peter Cock]
* Put unique prefix on contig names in --compliant mode [Peter Cock]

v1.8 - Nov 2013
* Added support for 'minced' CRISPR finder [Connor Skennerton]
* If --cpus > no. cores in system, reduce it down
* Improved message explaining CDS overlapping existing RNAs
* Defend against Bio::SearchIO bug for parsing hmmscan [Andrew Page]

v1.7.2 - 2 Nov 2013
* Fixed serious++ memory leak if --genes/--compliant used [Geoff Winsor]

v1.7.1 - 10 Oct 2013
* bug when --cpu 0 set [Lionel Guy]
* added two more entries to FAQ
* check for BioPerl >= 1.6.2 (not 1.6) for hmmer3 parsing [Bastien Chevreux]
* typo: "againt" => "against" [Dieter Bulach]
* cleanup product exception for 'rep' phage protein [Dieter Bulach]
* check for Parallel >= 20130422 bug https://savannah.gnu.org/bugs/?34241 [Nicola Soranzo]
* BUG: rnammer used -S bac no matter what --kingdom was [Harald Gruber-Vodicka]

v1.7 - 3 Sep 2013
* complete overhaul of method of using GNU parallel, twice as fast now!
* add support for my BARRNAP rRNA predictor software, will use before RNAmmer
* added --norrna flag to disable rRNA searching
* added --notrna flaf to disable tRNA searching
* add --plasmid option to enable /plasmid in 'source' tag
* add .txt output file summarising feature counts
* added name/version/author to --help output
* bugfix: parallel *.seq command line too long with >10^5 contigs [Jon Graf]
* updated kingdom uniprot databases
* filtered hypothetical/UPF/DUF proteins from all HMMs
* put TIGRFAMs before Cdd as COG names showing their age now
* bug: didn't check if <contigs.fa> was a readable, non-empty file
* option --moncontig renamed to --moncontiglen (backward compatible)
* default value for --mincontig set to 1 rather than 0
* exit if input fasta file doesn't contain any valid sequences
* added new binaries/{linux,darwin}, only used if not already installed

v1.6 - 7 Aug 2013
* fixed bug: 'infernal' exe doesn't exist, look for 'cmscan'
* added custom trusted E.coli --proteins file derived from EcoCyc
* added prokka-biocyc_to_fasta to make Prokka databases from Biocyc downloads
* leave contig names alone if --centre is NOT set
* make --centre '' the default
* fixed bug where --centre '' caused faulty /protein_id and no CDS in .gbk
* new --metagenome for working with fragmented genomes (sets prodigal --meta)
* new --compliant option to auto-set options for NCBI submission
* dont remove CDS overlapping ncRNA found via --rfam
* document the "tbl2asn" out of date problem!
* add --gnu option to parallel and require version XXX
* fixed problem where .gbk had no CDS features (added "-a s" to tbl2asn)
* added parallel and blastp to version checking
* enforcing minimum version of software to ensure compatibility
* signalp(--gram),makeblastdb(--proteins),cmscan(--rfam) are now optional installs
* added support to add platform binaries to binaries/{linux,darwin}/
* added prokka-build_kingdom_dbs to enable user updating of uniprot BLAST dbs
* added prokka-build_genus_dbs to enable user to generate new ones ???
* added support for binaries/linux and binaries/darwin to be included
* fixed bug: wrong detection of rfam index file

v1.5.2 - 14 Sep 2012
* fixed bug to handle -ve strand ncRNAs [Adam Caldwell]

v1.5.1 - 13 Sep 2012
* fixed bug parsing cmscan output [Adam Caldwell]
* added Infernal to documentation requirements
* record $USER in .log file

v1.5 - 9 Sep 2012
* fixed .faa output not honouring --gctable [Adam Caldwell]
* option --addgenes causes duplicate entries in the .ffn output
* removed --coverage option which wasn't being used/honoured
* changed /usr/bin/perl to /usr/bin/env perl
* added support for tmRNA (ssrA/10S) features from Aragorn
* added support for Infernal/cmscan 1.1 against Rfam
* added a "Bug" to docs
* consisified some warning messages
* added LICENSE for Rfam and Pfam
* removed redundant sprot fasta files: prokka/db/kingdom/*/sprot

v1.4.1 - 14 Aug 2012
* fixed bug which rejected --increment 1 default! [Simon Gladman]

v1.4 - 13 Aug 2012
* improved cleaning up of product names to (help) satisfy NCBI tests
* added LICENCE information for Swissprot, TIGRFams in doc/ [Emailed Uniprot]
* option to change locus_tag numbering increment [Roderick Felsheim]
* fixed bug where setting --gcode was over-ridden by --kingdom

v1.3 - 17 Jun 2012
* removed 'sprot' fasta files, can recover from indices with fastacmd/blastdbcmd
* removed Time::Seconds->walltime() as it is too new [Heikki Lehvaslaiho]
* checked for GNU parallel VS "parallel" from Debian moreutils pkg [Heikki Lehvaslaiho]
* fixed 'ribsomnal' / 'Ecoli.faa' typo in manual
* added citations for GNU Parallel and all other tools [Ole Tange]
* add /protein_id tags to CDS [NCBI]
* added --citation option to show publication reference
* improved manual w/ details on how to get RNAmmer working with old HMMER
* added "Installation" section to manual
* changed NCBI error suggestion file from *.val to *.err
* added prokka-make_tarball script to automate new releases

v1.2 -  28 Apr 2012
* fixed changelog 1.01 to 1.1 as released
* fixed bug so only newly created files are listed at end when --outdir='.'
* work-around for bug in Aragorn reporting 27000bp long tRNAs at contig ends
* wrote blog post about it at http://thegenomefactory.blogspot.com/
* announced on Google+, Twitter

v1.1 - 25 Apr 2012
* fixed accidental forcing lower case for strain instead of species
* use of genus-specific DBs is now default off, use --usegenus to enable it
* changed $URL
* altered COMMENT about prokka in .gbk file
* features weren't sorted by location before locus tag assignment
* identified bug in tbl2asn introducing space in /inference="COORDINATES: profile"
* added support to "un-collide" duplicate /gene names by adding _1 _2 etc
* added more documentation in doc/prokka-manual.html
* added --listdb option to show what databases are configured
* removed .gz versions of raw HMM databases, as not needed

v1.00 - 23 Apr 2012
* first public version
