Decompose explicit functional groups in SMILES strings

A lightweight, dependency-free SMILES graph parser identifies common metabolite functional groups and derives interpretable positive-ion, negative-ion, alkali-binding, and neutral-mass priors. The parser is intended for candidate prioritisation rather than pKa, proton-affinity, or quantum chemistry calculations. Stereochemistry is accepted but does not change the counts.

Usage

DeconvolveSMILES(
  smiles,
  backend = c("auto", "native"),
  strict = FALSE,
  workers = getOption("SpaMTP.smiles_workers", 1L)
)

Arguments

smiles: Character vector of SMILES strings.
backend: Currently "auto" and "native" both use SpaMTP's native parser. The argument reserves a stable extension point for optional chemistry toolkits.
strict: Stop when any SMILES cannot be parsed. When FALSE, invalid rows are returned with structure_valid = FALSE and missing features.
workers: Number of forked workers used for unique SMILES on platforms supporting parallel::mclapply(). The default is controlled by options(SpaMTP.smiles_workers = 1).

Value

A data frame with functional-group counts, ion-mode scores, and a human-readable structure_evidence field.

Examples

utils::str(formals(DeconvolveSMILES))
#> Dotted pair list of 4
#>  $ smiles : symbol 
#>  $ backend: language c("auto", "native")
#>  $ strict : logi FALSE
#>  $ workers: language getOption("SpaMTP.smiles_workers", 1L)