I'm working on a metabolite database where "labeled" compounds can contain isotopically labeled atoms in specific positions. When a new labeled compound is entered into the database, we'd like to be able to perform a simple validation on the submitted labeled isotope positions. E.g. there's a Carbon-13 at position 2 in a glucose compound.

The first suggestion by another developer was to check that the positions submitted are less than or equal to the number of atoms of the given type in the compound based on the formula. It's been a long time since my BS in Biochemistry, but I recalled that positions are not (necessarily?) relative to the given atom. The positions (as I recall) are based on multiple types of atoms and are numbered in order based on its weight and the weight of atoms it's bonded to, though I don't remeber the specific rules.

I've been googling for awhile now and flipping through my old chemistry books, but I have yet to turn up those rules that govern:

  1. What atoms receive a position (e.g. carbon and nitrogen atoms are numbered 1-9 in purine but oxygen doesn't get a position in glucose)
  2. What atom gets position 1, or any other position for that matter

For now, all I want to do is determine the total number of numbered positions so that if an isotope position is submitted outside of the range, we can throw an error and say that position is invalid. Aside from having more C13 positions than the number of carbons in the compound, that's all our requirements currently call for. (While it would be nice to know, we don't have a requirement for ensuring the given position actually is supposed to be a carbon.)

And while I'm at it, are there different atom position/numbering schemes in use currently?

  • $\begingroup$ Could it be as simple as summing all the atoms from the formula that excludes oxygen and hydrogen for example - could that be thwarted by some edge cases, like diethyl-ether, where there's an oxygen in the middle, or other un-numbered "side-groups"? $\endgroup$
    – hepcat72
    Jun 3 at 20:45
  • 1
    $\begingroup$ I'm not sure I get you, why not just use the number of atoms of the element? Using the number of magnetically equivalent atoms would be better, but I don't see a point, if you just want to cut down errors. $\endgroup$
    – Mithoron
    Jun 3 at 21:44
  • $\begingroup$ @mithoron. The point is we don't want erroneous data in our database. We want validated/correct data. Not sure I'm following you, though I could take the total and subtract all the hydrogen atoms as the most basic heuristic, but the point is, I would like to know the rules enough to be able to make an intelligent check. Like, aside from hydrogen, are there any atoms that don't get a position? Do hydrogen atoms not get a position or do they get positions, but their positions are always last, numerically? Basically, what are the numbering rules? I'd like to do better than to low bar of all atoms $\endgroup$
    – hepcat72
    Jun 5 at 2:44
  • $\begingroup$ Wait, is there a miscommunication? I'm not counting protons/neutrons of an "element". I'm counting occurrences of elements (I.e. atoms) in a compound. $\endgroup$
    – hepcat72
    Jun 5 at 2:46


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.