Incorrect Iron Mass

[hbckleikamp]

Hi,
I am using the Sirius decomp tool to predict molecular formulas, but I noticed the mass you are using for Iron is incorrect.
There is a ~2 Da mass error in the predictions that contain an iron atom. I think you are using the mass of 54Fe : 53.93960899 instead of 56Fe: 55.93493633, which is the standard, monoisotopic mass.

[mfleisch]

Hey,
thanks for reporting. Can you provide some example to reproduce this problem (I know it has been a while since you reported this).

"Fe": {
        "name": "Iron",
        "valence": 1,
        "average": 55.84515022389856,
        "isotopes": {
            "mass": [53.939615, 55.934942, 56.935399, 57.933280],
            "intmass": [54, 56, 57, 58],
            "abundance": [0.05845, 0.91754, 0.021191, 0.002819]
        }

I checked our periodic table and it seems to be correct. so the problem seems to be caused by something else....

[hbckleikamp]

Hey, no worries, yes all of this element info is correct.

I think it's more of a definition thing, because the monoisotopic mass of iron is 55.934942, but when Sirius does MFP it uses the lowest mass isotope 53.939615 instead of the most common monoisotopic isotope. I think this lead could to some errors in the isotopic composition, as Fe53 Fe55 should be seen as -2 & 0 and not 0 & +2 isotopes. Not many people use these kinds of elements in their predictions though.

[mfleisch]

Hmm,
expected behaviour would be to use the most common isotope is used. Do you have some iron containing compound spectrum lying around we can use for testing?

[JethroHemmann]

I am currently also trying to use SIRIUS 6 for an iron-containing compound. It fails to find the correct formula. (Fe was selected as additional element)

Attached you find an MGF with the MS/MS of a precursor that has iron bound. The most intense peak in the pattern was fragemented, thus the compound will contain Fe56 (55.9 Da) and not Fe54 (53.9 Da).

The correct formula would be: C28H44N9O13Fe

Fe-compound.txt

[JethroHemmann]

Has there been any updates regarding the annotation of iron-bound compounds?

[kaibioinfo]

The problem is that we always assume lowest-mass-isotopes for the fragmentation pattern analysis. The main reason is, that the fragmentation pattern for other isotopes can become very complex and contain isotope pattern itself. At least this is the case for elements like carbon, oxygen or hydrogen.
For Iron, however, it is very unlikely to observe more than one iron in a compound. So we could just make an extra rule that allows for the Fe56 instead of Fe54. However, that would be hardcoded - something we usually avoid. Given that iron compounds are of high research interest we might add such a rule in future updates, though.