Annotate MS2 precursor peaks using a compound database

Annotate object@advancedAna$MS2_Precursor (from MSdev_get_peak_table_from_spectra) with CompDb MS1 candidates and MS2 spectral scores. Same CompDb engine as MSdev_annotation, but the targets are spectra-derived precursor peaks, not xcms MS1 features. Isotope-pattern scoring is skipped (no MS1 intensity matrix). Requires ms2_id on the peak table linking to MS2_Spectra sp_id / spectraNames.

Usage

MSdev_annotation_MS2_Precursor(
  object,
  cpdb_path = "c:/Users/91879/OneDrive/Code/R/data/MSDB/CompoundDB/CompoundDB.sqlite",
  ppm = 10,
  weight_mz = 0.2,
  weight_ms2 = 0.8,
  BPPARAM = BiocParallel::SnowParam(workers = max(1L, floor(parallel::detectCores()/3)),
    progressbar = TRUE),
  ...
)

Arguments

object

MSdev object with advancedAna$MS2_Precursor

cpdb_path

path to CompoundDb SQLite database

ppm

m/z tolerance in parts per million for candidate matching

weight_mz

weight for m/z error score (default 0.2)

weight_ms2

weight for MS2 similarity score (default 0.8)

BPPARAM

BiocParallel backend for MS2 scoring. Default is a SnowParam with floor(detectCores() / 3) workers. Polarities are processed one after the other.

...

additional arguments passed to annotation helpers

Value

MSdev object with annotated advancedAna$MS2_Precursor

Details

MS2_Precursor is not from xcms peak picking. It is built by grouping MS2 spectra by precursor m/z (ppm) and RT gap (rt_tol) into rows such as MS2P000001, each with mzmed/rtmed and ms2_id pointing at those spectra.

For each polarity the function:

1. matches precursor mzmed to CompDb adduct m/z (fdf_get_ms1_candidate);

2. scores experimental MS2 vs CompDb reference spectra (fdf_get_ms2_score, ndotproduct);

3. sets score.isopattern to zeros (no MS1 intensity matrix);

4. picks the best candidate with default weights weight_mz = 0.2, weight_ms2 = 0.8, weight_isopattern = 0.

Results (compound_id, adduct, score, CompDb name/formula/smiles, ...) are written back to advancedAna$MS2_Precursor. projectInfo$CompoundDB_path is also set. If MS2_Precursor is missing, a message asks to run MSdev_get_peak_table_from_spectra first and the object is returned unchanged.

Difference from MSdev_annotation:

	MSdev_annotation	MSdev_annotation_MS2_Precursor
What	xcms MS1 features (PositiveMS1 / NegativeMS1)	MS2 precursor groups (advancedAna$MS2_Precursor)
Written to	featureDefinitions on the xcms object	advancedAna$MS2_Precursor
MS2 link	ms2_id from MSdev_assign_MS2	ms2_id from grouping the MS2 spectra themselves
Isotope score	optional (calc_isopattern_score)	never
Default weights	mz 0.1 / MS2 0.7 / iso 0.2	mz 0.2 / MS2 0.8 / iso 0
Helpers	xcms_get_feature_*	fdf_get_* (same logic on a data.frame)

The two functions do not overwrite each other.

See also

MSdev_annotation, MSdev_get_peak_table_from_spectra