More precursor ion options and candidate limits

[sneumann]

Hi, I got a question:

1. Ist es möglich, dort in der Liste mit den "parent ions" auch Dimere ([2M+H]+, [2M-H]-) zu implementieren? Es sind zwar [M-2H]- implementiert, aber keine Dimere. Wir haben manchmal Stoffe, die hauptsächlich als Dimere gefunden werden und da wäre das sehr hilfreich.

Currently additional precursor ion types in the calculator are not readily available. Of course, you can calculate the neutral mass manually and add for the candidate search. Some external software (including Compass3D) can create a Landing URL, which looks
like https://msbi.ipb-halle.de/MetFrag/landing.xhtml?FragmentPeakMatchAbsoluteMassDeviation=0.001&FragmentPeakMatchRelativeMassDeviation=5&DatabaseSearchRelativeMassDeviation=5&PeakList=57.06984_29;61.97913_88;68.02435_999;71.0604_228;79.00582_486;90.01058_17;96.05565_432;104.00108_999;110.04612_66;128.05657_6;132.03232_422;138.0775_129;146.02289_393;172.0386_4;174.05425_999;230.11679_999&IonizedPrecursorMass=230.1162&MetFragDatabaseType=PubChem
and includes all parameters pre-filled.

If someone wanted to create a pull request, one of the places to look would be https://github.com/ipb-halle/MetFragRelaunched/blob/c57f9d2b406350b2357ce9f7ce42a286cefcca13/MetFragLib/src/main/java/de/ipbhalle/metfraglib/parameter/Constants.java#L138 but other places would need changing as well.

2. Manchmal sind wir bei "retrieve candidates" schon über der maximal erlaubten Anzahl an Kandidaten und können dann nicht weiterarbeiten. Wir geben schon die Summenformel vor und wissen nicht wie wir die Liste noch weiter einschränken sollten. Gibt es dafür eine Lösung?

Indeed, that was because we had issues with upstream databases blocking denial-of-service attacks.
The command line version has no such limits.

You could attempt "Candidate Filter & Score Settings", left column with candidate filter.
If you already have the molecular formula, then only substructure in/exclusion or suspect lists can help.

Also (and what a timing!) the Candidate Database drop-down includes "local databases",
which are collections of compounds, many of which contributed by Emma Schymanski.
The Pubchem Tier0 / Tier1 are subsets of PubChem which have some annotations (literature, patents, function, ...), and hence are likely to be more relevant than some of the synthetic stuff one might encounter in PubChem https://zenodo.org/record/3611238

Also of interest could be a list of natural products collected in the team of Chris Steinbeck:
https://www.preprints.org/manuscript/201912.0332

If one want to deploy a local instance of MetFrag, this would be the place to change:
/blob/0e69e62d3c2428367b22bc7bf2c290afe1d8e402/MetFragWeb/src/main/java/de/ipbhalle/metfragweb/container/BeanSettingsContainer.java#L202

Yours, Steffen