Furthermore, the DOM is rich in heteroatoms (e.g., N, S, P, and Cl) and its molecular formulas are present in the lipid-, protein-, carbohydrate-, and lignin-like regions of van Krevelen diagrams.
These molecular formulas are the nodes of a directed acyclic graph (DAG) called fragmentation graph.
The molecular formulas are given in Figure 2. Figure 2 Molecular structures of PTEA 11K - b -PAM 30K, PDADMAC, PEI, and PAH.
For the prior (P_{mathrm{root}}(r)) of the root r we use the molecular formula r, and the fact that certain molecular formulas are observed more often in molecular databases [20].
Furthermore, prior probabilities are chosen conservatively, so that "unlikely" molecular formulas are penalized only slightly: if there is sufficient MS/MS data, SIRIUS 3 will "overrule" these priors, returning a molecular formula that violates one or even several prior assumptions.
As a theoretical example, consider masses below 15 Da: in this mass region, reasonable molecular formulas are H, (hbox {H}_{2},) CH and (hbox {CH}_{2}.) To this end, all peaks with other masses must be noise.
For CASMI, an internal database was used to prioritize existing formulas from large chemical databases over less common formulas and the top 5 molecular formulas were regarded for structure queries.
Molecular formulas were predicted using Bruker SmartFormula algorithm, which uses both exact mass and isotopic patterns.
The same molecular formulas were also searched against publicly available Human Metabolome Database HMDBB, http://hmdb.ca) [ 23].
These molecular formulas were matched against available databases, namely Metlin, Human Metabolome Database, and Scifinder.
These molecular formulas were searched against the Metlin compound database (Metlin_AMRT_PCDL.cdb) using the ID browser application of MPP.
