This version upgrade has been a long time in the making. Sorry it has taken so long. I kept meaning to release sooner but then new feature requests or bugs made me delay. But This means you get a very full set of new features and bug fixes.
Of special note, We have switched the BAM reading code from BamTools to samtools/htslib so you no longer need to install BamTools to compile ZENBU. Also there is a new auto-build script to easy the installation on debian and ubuntu systems. Please check out the README_AUTOBUILD.
2.10.0
Added easy build script zenbu_build_debian.sh. See README_AUTOBUILD for details
Added ranksum visualization option to the infopanel when in mann-whitney statistical mode
Added option to show error bars are either standard error or standard deviation when in grouping statistical mode
Rebuilt the BAM reading code around samtools htslib library function calls. Previously was using the BamTools c++ library. samtools introduced an incompatibilty with version 1.2. Sometimes BAM files built with samtools1.2 would cause the BamTools library to crash. So we decided to rebuild the code around the samtools library.
Fixed bug in the mann-whitney ranksum calculation since we want to preserve the directionality of the ranking in the z-score
Fix bug in BAM paired-end parsing code. Now correctly assigns the strand for the mated-pairs. Still does not parse the edge/link inforation, but this will happen in a future release.
Added processing module FlipStrand
Added "unshare" feature. Allows a user to unlink/unshare a data-set from a collaboration.
Added key-specific metadata searching to entire system. Two new syntax are added
key:=value (eg UBERON:=0000061) is an exact match for this specific metadata
key~=value (eg UBERON:desc~=bone) is an exact key match and partial value match (like bone)
Added hover-over info panels from mdata-grouping mode. click shows detail panel with all individual experiments in the metadata key-group and hover-over the experiments shows for info and click on experiment show full experiment metadata. Improved formatting of these panels. Removed "keyword" from the experiment detailed metadata display.
Added new visualization mode into the genome browser called "split-signal". For each experiment / or experiment grouping in the linked expression-panel, a separate sub track is displayed in the signal-histogram form. All subtracks are put on a common scalling so that if an experiment/grouping is lower signal, than the max height of the subtrack is proportially lower. Changing the sort order in the linked expression panel changes the order of the sub-tracks in the split-view. If the linked panel is in a sort mode of signal then the sub-tracks will also dynamically resort themselves as a selection is made. The track-height parameter sets the height of each sub-track and the total height of the track is the track-height times the number of experiments/groupings. Although this works for a non-grouped tracks, it is most useful when the tracks experiments are grouped into clean categories or there are only a few experiments pooled together in the track.
Added remote capability to the region-sequence web service.
Added strand and reverse-complement capability to the region-sequence web service and interface. Sequence on features uses the strand of the feature to determine if it rev-comp. sequence of selection uses the global orientation of the view. Strandless features uses the global orientation.
Added search= to URL for genome browser to allow recentering on an exact match search term. Can be anything searchable in the view from entrez genes to gencodeIDs. eg: fantom.gsc.riken.jp/zenbu/gLyphs/#config=b-IMGb5lG53ntH8qgNeChB;search=egr1
Added active_track_exp_filter= to genome browser URL parsing. A very special mode where it performs this experiment-filter-search on the selected track at launch time. Allows for a common view of pooled data to be easily hyperlinked into from an external application where a specific filter of experiments is required.
Added keep_active_experiments_on_top option to the linked experiment-expression panel. Keeps the selected experiments at the top of the panel (with selected sort order) and then the deselected will follow (also in choosen sort order)
Added a "global no cache" for forcing direct queries. options is not saved in view configuration. mostly used for debugging.
Converted Collaborations to use sqlite databases for their registries. Originally used sqlite, then switched to mysql, but now realized that many mysql registries cause problems exhausting available network ports. So switched back to sqlite and will continue to remove mysql from the system in the future.
fixed small bug in tracking line where it wouldn't click to base 1 when zoomed in at a 1bp level
fixed display of sequence length of a selection
Improved the performance of feature-name-searching within ZDXdb (eg GFF loaded data with name searching). Now the system finds an intersection of potential znodes blocks based on the mdkeys.
Improved GFF uploader so that formating errors and parser failures are caught and detailed information is passed back to the user so they can fix their file formatting.
Improved performance of GFF uploader when dealing with draft genomes of 1000s of chromosome/scaffolds
Fixed NCBI gene loader. NCBI changed the output format for the entrez gene webservices (http://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi). Had to rebuild scripts/eedb_sync_entrezgene.pl XML parsing
Fixed bug when uploading BED files with score-count or single-count. The automatic normalization ("_pm" datatype) was not being triggered correctly. This will be fixed for all new uploaded. For files which were previously uploaded, but missing the _pm datatype, we are considering a "repair" tool.
Fixed bug in the UniqueFeature code when in "strandless" mode. In some situations it would not properly merge the different strands and return stranded features
Fixed some bugs in the zenbu_upload -mdedit system.
Fixed small bug in eedb_chromChunkTool.pl to now allow files with extensions .fa .fa.gz .fasta .fasta.gz
Fixed bugs in the data explorer DEX URL parsing and init for collab= and search=
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Please note that the zenbu_build_debian.sh script is for new server installation and not for server upgrade. I can make another version of the script for upgrading if it is requested.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
uploaded patch version ZENBU_2.11.0.b.tar.gz today. Forgot to change the zenbu_build_debian.sh to 2.11.0 and forgot to include the /build_support directory. This autobuild script will now work in 2.11.0.b
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
This version upgrade has been a long time in the making. Sorry it has taken so long. I kept meaning to release sooner but then new feature requests or bugs made me delay. But This means you get a very full set of new features and bug fixes.
Of special note, We have switched the BAM reading code from BamTools to samtools/htslib so you no longer need to install BamTools to compile ZENBU. Also there is a new auto-build script to easy the installation on debian and ubuntu systems. Please check out the README_AUTOBUILD.
2.10.0
key:=value (eg UBERON:=0000061) is an exact match for this specific metadata
key~=value (eg UBERON:desc~=bone) is an exact key match and partial value match (like bone)
Please note that the zenbu_build_debian.sh script is for new server installation and not for server upgrade. I can make another version of the script for upgrading if it is requested.
uploaded patch version ZENBU_2.11.0.b.tar.gz today. Forgot to change the zenbu_build_debian.sh to 2.11.0 and forgot to include the /build_support directory. This autobuild script will now work in 2.11.0.b