Compound Identifiers and Isomers

Here we outline the three different compound identifiers that are used and their distinctions. We also describe certain methods used in the config command.

1. Ion Formulas

Mass spectrometry begins with the ionization of molecules within a sample before they are passed to the detector. Hence, what is being detected is actually an ion. So, many peak lists might rightfully give their peak identifiers as ion formulas.

This is why we accept ion formula peak lists as inputs to the config command. We translate the ion formulas to molecular formulas by subtracting an H using regular expressions, and then search our database for the most common isomer of that molecular formula.

2. Molecular Formulas

While ions are what is detected by a mass spectometer, neutral molecules are usually what is present in the sample. We made the decision to use molecular formulas, not ions, as our primary identifier in our database used to do substructure matching.

If only a molecular formula is given in the input to the config command, then we search our database for the most common isomer of that molecular formula.

3. SMILES

SMILES are identifiers that describe the structure of a chemical species. This makes them more specific than either of the above identifiers, as a single formula might have many different isomers with unique structures (and therefore unique substructures!).

The most foolproof way to create a peak list configuration file is by providing a column with SMILES strings for each peak / compound. This way there is no ambiguity in the database search.