search

ipb-halle / MetFragRelaunched

Relaunch of the initial MetFrag project.
http://ipb-halle.github.io/MetFrag/
17 stars 16 forks source link

different OfflineMetFusionScore values #90

Closed gilcu3 closed 2 years ago

gilcu3 commented 2 years ago

Environment: Linux java-11-jdk

parameters file: parameters34cb6d1cb2af2f.txt spectrum file: spectrum34cb6d3ae3d66b.txt local database: PubChemLite_31Oct2020_exposomics

With metfragCL 2.4.5 the output is: results34cb6d196b0dd-metfrag2.4.5.csv With metfragCL 2.4.7 the output is: results34cb6d196b0dd-metfrag2.4.7.csv

Using version 2.4.7 the column OfflineMetFusionScore is filled with 0s, while it should not be, at least according to version 2.4.5.

I got the same problem for column OfflineIndividualMoNAScore in another experiment.

schymane commented 2 years ago

Likely related to the MoNA parts of https://github.com/ipb-halle/MetFrag/issues/22 - MoNA is also broken on the newer web versions...

sneumann commented 2 years ago

We might have found the culprit in https://github.com/ipb-halle/MetFragRelaunched/commit/1a26a9ac29322525e2f72016af327a3915905032 and @meier-rene is on it to fix. A continuous integration performance regression test by @c-ruttkies is in the works to catch things like this. Yours, Steffen

schymane commented 2 years ago

OK great, because we still have our "back-up" running (as we need it atm), but it costs $$$ every day to keep running ... ;-)

meier-rene commented 2 years ago

Fixed and checked or identical results. Thank @gilcu3 for your detailed report with example to test. I will release a new version 2.4.8 which includes this fix.