If you run dgenies in webserver mode, database schema has changed.
The column batch_type
was changed to runner_type
in job
table.
You can update the database with following sql command:
ALTER TABLE job RENAME COLUMN batch_type TO runner_type;
This change was also reflected for cluster configuration in application.properties
, where batch_system_type
was renamed to runner_type
.
For users:
- Update batch mode:
- Local files upload
- Batch content checked in web interface
- Single upload per file
- Batch file format has changed, in particular, option keys have been change to be more comprehensible. Please read the doc.
- Dotplot:
- Files generated through "Query Fasta" and "Query assembled as reference" are updated if new sorting and contig reversal are done.
- Files linked in email are stable until the job is cleanned. They dont change if sort and contig reversal are updated.
- You can reset the sort (usefull when playing with manual contig reversal)
- "Query assembled as reference" file is now compressed in webserver mode.
- Job logs are now available to download and are included in backup file.
- Update user interface to
bootstrap 4.6
For sysadmin:
- Tools config file format (
tools.yaml
) has changed. Please update it if you use a custom one. - Basic app logging.
- Configuration files:
- In
application.properties
,batch_system_type
was renamed torunner_type
in order to avoid confusion with batch jobs intoduced with dgenies 1.4. - Cluster
walltime
can be set in the config file. - In database, the column
batch_type
was changed torunner_type
injob
table. - It is possible to set config files directly when running dgenies with
--config
and--tools-config
options. Please be careful to use the correct config files when managing a running instance of dgenies. - It is now possible to use a the option
--flask-config
in order to set some flask options. You can use it to set some email server parameters. - For server mode, the number of download sessions in parallel can be now set with
max_download_sessions
(default is alway5
). Session delays can also be modified now. - Reinforce tar file checking
- Drop Flask<2
- In
- When sorted, "Association table", "Query Fasta" and "Query assembled as reference" files match dotplot. Before, only the last reversed contig on dotplot was token in account.
- Fix some bugs with Flask>=2
- Add compatibility with Flask 2.2
- Some gzipped files were uncompressed by browser when downloaded.
- Backup file name was incomplete.
- Add batch mode. This early version only allows to use urls as file inputs. #28
- Add definition page to explain how some things like Identity, Best matching chromosomes are computed.
- Relax Jinja and Flask requirements up to last versions
- Some options were display to right side of page
- Add some anonymization strategies
- Fix some Jinja2 requirement constraints
- Fix exported query fasta file that was not compressed when asked.
- Add
repeatedness
option with minimap2. This option needs to upgrade the database schema with following command:ALTER TABLE job ADD COLUMN options VARCHAR(127) NULL;
- Backup archive is now a
tar.gz
file. Oldtar
backup archive are still supported - Upgrade embedded minimap2 to the latest available version (2.24)
- GDPR compliant (webserver mode):
- Cookie wall option
- Ability to set legal stuff pages
- Analytics is now anonymous, an option allows to restore the previous behavior
- Python 3.10 compatibility
- Add a command to clean analytics DB from dgenies (
dgenies clear -a [--max-age <age>]
) - Check if PAF file is correctly formatted
- Performance improvement
- Expose
max_nb_lines
parameter in configuration file. - Tools can have a label now (in
tools.yaml
). - Speedup compressed file operations by using
xopen
library - Fix upload form. Form is now correctly reset when a field is missing or erroneous.
- Fix filename collision. Uploading a query file and a target file with the same filename now works correctly
- Fix wrong RAM usage displayed in gallery
- Correct default
slurm
parameters to match last version of slurm-drmaa (1.1.3) - Remove local scheduler pid file when stopping it
- Update documentation:
- Explain how similarity/identity measure is computed
- Add link in D-Genies to what is expected in backup archive
- Wiki
- Cookbooks
- Additional information for developers
- Update python requirements
Bugfix release
- Dot plot have now a mouse cursor centered zoom
- Now includes mashmap (v2.0), a faster aligner for high identity genomes, as an alternative to minimap2 (Linux only)
- Download a HTML page which permit to show an interactive dot plot offline (with the summary)
- Display match coordinates by moving the mouse cursor over the match line
- Upgrade embedded minimap2 to the latest available version (2.11)
- Add help messages on the interface
- Add the ability to add a message to the run form through command line (designed for maintenance for example)
- Several bug fixes
- Fix bug with parse of MAF files
- Fix incompatibility with pip >= 10.0
- Improve run form: add support for uploading an alignment file (PAF or MAF by default) instead of fasta files, if alignment file is already available
- Add export of a job into a backup, which can be re-uploaded with the run form
- Ease the integration of new tools or new file formats
- Allow to install as simple user on Unix systems
- Export of the dot plot is now done including the zoom factor
- Add new color palettes for dot plot
- Add analytics logs (anonymous) which permits to know which size of genomes are ploted (can be disabled)
- Improve documentation
- Minor bug fixes
Bug fix release.
- Fix bugs
- Disable export of fasta files for gallery items
First stable release