Skip to contents

Calculate Isotopic Pattern Similarity Scores

Source: R/formula_engine.R
scoreIsotopes.Rd

Computes modern spectral similarity metrics including Cosine similarity, intensity-weighted dot product, and log-likelihood between observed mass spectral peaks and theoretical isotopic distributions.

Usage

scoreIsotopes(
  obs_mz,
  obs_int,
  theo_mz,
  theo_int,
  tolerance = 0.005,
  ppm = FALSE
)

Arguments

obs_mz

Numeric vector of observed m/z values.

obs_int

Numeric vector of observed intensities.

theo_mz

Numeric vector of theoretical isotopic m/z values.

theo_int

Numeric vector of theoretical isotopic abundances.

tolerance

Numeric mass tolerance for peak alignment (default: 0.005).

ppm

Logical indicating whether tolerance is in ppm (default: FALSE).

Value

A list containing:

cosine

Cosine similarity (unweighted dot product of normalized spectra).

weighted_cosine

Mass- and intensity-weighted dot product (MassBank/NIST style).

log_likelihood

Multinomial log-likelihood of observed peak distribution.

matched_peaks

Integer count of successfully matched peaks.

Examples

obs_m <- c(180.0634, 181.0668)
obs_i <- c(100, 6.5)
theo_m <- c(180.0634, 181.0668)
theo_i <- c(1.0, 0.066)
scoreIsotopes(obs_m, obs_i, theo_m, theo_i)
#> $cosine
#> [1] 0.9999995
#> 
#> $weighted_cosine
#> [1] 0.9999983
#> 
#> $log_likelihood
#> [1] -0.2298068
#> 
#> $matched_peaks
#> [1] 2
#>