BaMMserver v1.4.0

We are happy to announce version 1.4.0 of the webserver which will be online starting from August 9th. 🎉

New features

• we now offer a drosophila dataset based on the ModERN dataset. (#132)

Bug fixes

• fixed a couple of small bugs that lead to server crashes (#128, #129, #130)

Under-the-hood improvements

• cleaned up dependencies and updated django to 2.0.8 (#126)

BaMMserver v1.3.1

We are happy to announce version 1.3.1 of the webserver which will be online starting from July 18th. 🎉

Bug fixes

• fdr could sometimes crash due to invalid memory accesses (#125)
• invalid uuids can no longer crash job search (#122)

Under-the-hood improvements

• updated docker image base to alpine 3.8 (#126)

BaMMserver v1.3.0

We are happy to announce version 1.3.0 of the webserver which will be online starting from July 2nd. 🎉

New features

• bamm files and background fasta files are now validated before job submission

Bug fixes

• distribution plots are now created for all motifs in the refine workflow
• meme validator now handles files that cannot be decoded to utf-8
• non-uuids in job search by id no longer crashes the server
• fixed problems when calculating joint probabilities from conditional probabilities

Under-the-hood improvements

• backups are now cleaned up as expected
• extra config allows email logger to report errors in requests

BaMMserver v1.2.1

We are happy to announce version 1.2.1 of the webserver which will be online starting from June 21st. 🎉

Bug fixes

• the server now properly reports back errors to the user and is not stuck in the "still running" status
• BaMM files with dots in the filename do not crash MMcompare anymore
• fixed a problem where uploaded BaMM files could crash BaMMscan

Under-the-hood improvements

• the server now supports regression testing to avoid breaking again what we fixed once
• we changed the docker base image to alpine and reduced image size in half :-)

BaMMserver v1.2

We are happy to announce version 1.2 of the webserver which will be online starting from June 13th 9am UTC. 🎉

New features

• the "Find my job" page now explains visibility of jobs and allows to download the full list of jobs available on the server with the new "Download as CSV button"
• bamm output page now ranks motifs by occurrence * motif AvRec - this way the first motif should be the motif that performs best weighted by occurences. No longer should very rare motifs appear to be best performing.

Bug fixes

• fixed problems where filtering short sequences could crash the FDR evaluation
• fixed problems where dots in file names were able to crash mmcompare

Under-the-hood improvements

• improved process management to ensure proper warm shutdown without job loss

BaMMserver v1.1

We are happy to announce version 1.1 of the webserver which will be online starting from June 1st 2-4pm UTC. 🎉

New features

• When all fasta headers are of the format <seqid>:<start>-<end> (e.g. chr1:100-250), a bed file of occurrences is generated automatically and a new button allows to visualize the occurrences in the UCSC genome browser.
• We now generate negative sequences with a sequence specific background model, making the performance evaluations more robust when the sequence composition between sequences is very inhomogeneous (e.g a substantial part of sequences coming from CpG islands).

Bug fixes

• very short sequences can no longer crash BaMMmotif
• fixed a problem with the HOCOMOCO databases leading to a server crash
• UI: fixed broken select-all checkbox in manual seeding workflow

Under-the-hood improvements

• cronjob with daily backups now automatically schedules with celery
• added celery cronjob for cleaning up expired jobs