Adduct Mass Calculation From Smiles

adduct-mass-calculation-from-smiles

Summary

Compute expected m/z values for metabolite adduct ions given a SMILES string and a derivatizing matrix profile (e.g., FMP-10). This skill automates the prediction of non-standard ionization products beyond [M+H]+ and [M-H]−, enabling high-throughput metabolite identification in mass spectrometry imaging.

When to use

When you have a metabolite SMILES structure and need to predict which adduct ions will appear in a mass spectrum acquired with a chemical derivatizing matrix. Use this skill especially when the derivatizing matrix produces ions other than common [M+H]+ in positive mode or [M-H]− in negative mode, such as with FMP-10 or other custom matrices. Apply it during the feature-to-compound matching stage of metabolite identification in mass spectrometry imaging workflows.

When NOT to use

• Input metabolite structure is already fragmented or is an MS/MS product ion rather than an intact molecule—use MS/MS spectral matching instead.
• The derivatizing matrix is not characterized or no adduct ruleset is available; attempting to guess the matrix profile will produce incorrect predictions.
• You are working with raw mass spectrometry imaging data and need to identify all peaks simultaneously; use this skill as part of a larger feature-matching pipeline, not in isolation.

Inputs

• metabolite SMILES string
• derivatizing matrix identifier or ruleset (e.g., 'FMP-10' or custom matrix configuration)
• optional: ionization mode ('positive' or 'negative')

Outputs

• table of predicted adduct ions with formula, mass shift (Δm), and expected m/z values
• ranked list of expected m/z values for the metabolite under the specified matrix conditions

How to apply

Parse the input metabolite SMILES string using RDKit to construct and validate the molecular graph structure. Load the matrix-specific adduct ionization ruleset—either from built-in reference data (e.g., FMP-10 Nature Methods profiles) or from a user-defined matrix configuration. For each adduct rule in the matrix profile, apply RDKit's molecular weight calculator to compute the expected m/z by accounting for the derivatizing matrix modification mass, the adduct formula (e.g., +H, +Na, +NH4), and the resulting ionization charge state. Return a ranked table of predicted adduct m/z values with their corresponding formulas and mass shifts. Validate the outputs by comparing against reference adduct masses from published datasets or experimental standards to confirm the matrix-specific modifications are correctly applied.

Related tools

• RDKit (molecular graph construction, structure validation, and molecular weight calculation from SMILES) — https://www.rdkit.org/
• Met-ID (complete metabolite identification framework integrating adduct mass prediction with MS/MS spectral matching and derivatizing matrix support) — https://github.com/pbjarterot/Met-ID

Evaluation signals

• Predicted m/z values match reference adduct masses from published FMP-10 or derivatizing matrix datasets within the expected mass tolerance (e.g., <5 ppm for high-resolution instruments).
• All expected adduct ions for the matrix are represented in the output table (no missing adducts for the given ionization mode).
• Molecular weight calculation is internally consistent: computed m/z for each adduct obeys (molecular_weight + matrix_modification + adduct_mass) / charge_state.
• SMILES parsing completes without errors and produces a valid molecular graph; invalid or malformed SMILES are rejected with clear error messages.
• Output can be directly compared to observed mass spectrometry peaks and shows reasonable agreement in rank order and spacing, validating the matrix ruleset application.

Limitations

• The skill requires a pre-defined or user-provided matrix ruleset; without accurate adduct ionization rules for the derivatizing matrix, predictions will be incorrect.
• Metabolite structural ambiguity from SMILES (e.g., stereoisomers, tautomers) does not affect m/z calculation but may complicate downstream compound identification.
• The skill predicts only the primary and common secondary adducts; multiply-charged ions, dimers, or rare fragmentation adducts are not automatically enumerated.
• Matrix-specific behavior such as salt adducts ([M+Na]+, [M+K]+) must be explicitly included in the matrix ruleset; default ionization rules may not capture all in-source modifications.
• Derivatizing matrix modification masses and adduct rules must be kept current; outdated or incorrect ruleset definitions lead to systematic prediction errors across all metabolites.

Evidence

• [readme] Met-ID has a particular focus on derivatizing matrices leading to other ions than the common [M+H]+ in positive mode and [M-H]- in negative mode: "Met-ID has a particular focus on derivatizing matrices leading to other ions than the common [M+H]+ in positive mode and [M-H]- in negative mode"
• [other] Parse input metabolite SMILES and matrix identifier using RDKit to construct the molecular graph and validate structure: "Parse input metabolite SMILES and matrix identifier using RDKit to construct the molecular graph and validate structure"
• [other] For each adduct rule in the matrix profile, apply RDKit molecular weight calculation to compute the expected m/z accounting for the derivatizing matrix modification and ionization state: "For each adduct rule in the matrix profile, apply RDKit molecular weight calculation to compute the expected m/z accounting for the derivatizing matrix modification and ionization state"
• [other] Return a table of predicted adduct ions with their formulas, mass shifts, and expected m/z values, and compare against reference FMP-10 adduct masses from the published Nature Methods dataset: "Return a table of predicted adduct ions with their formulas, mass shifts, and expected m/z values, and compare against reference FMP-10 adduct masses"
• [readme] Met-ID is extendable to use any derivatizing matrix with the tools to do local version changes right from inside the software: "Met-ID is extendable to use any derivatizing matrix with the tools to do local version changes right from inside the software"