Estimate chemical formula

Estimate chemical formula

estimate_chemical_formula(
  query_mass,
  isotope_summary,
  penalization = "none",
  penalization_multiplier = 0.001,
  timeout = 10
)

Arguments

query_mass

amu of an unknown specie

isotope_summary

isotopes that the unknown specie could be constructed with:

• label

• amu

• count_lb (optional)

• count_ub (optional)

• prior_weight (optional)

If penalization = "prior", a variable named "prior" must be specified (larger weights are bad)".

penalization

the method of penalizing the optimization

• none: minimize the absolute difference between predicted mass and "query_mass"

• count: none + minimize sum of absolute atom counts

• prior: none + minimize a weighted sum of absolute atom counts

penalization_multiplier

multiplier for the penalization functions within the cost function

timeout

seconds until a timeout occurs

Value

a list containing a summary of fit (absolute mass difference and penalization cost (if penalization != "none")) and isotope_summary with an added variable

Examples

library(dplyr)

probable_isotopes <- clamr::isotope_summaries %>%
  dplyr::inner_join(
    clamr::elemental_frequency_summary %>%
      dplyr::filter(human_body > 0.0001) %>%
      dplyr::select(element, human_body),
    by = "element"
  ) %>%
  dplyr::mutate(natural_freq = isotope_freq * human_body) %>%
  dplyr::filter(natural_freq > 0.0001) %>%
  dplyr::mutate(
    count_lb = ifelse(element == "H", -8, -2),
    count_lb = ifelse(element %in% c("C", "O"), -4, count_lb)
  ) %>%
  dplyr::mutate(count_ub = -1 * count_lb) %>%
  dplyr::rowwise() %>%
  dplyr::mutate(label = paste0(element, "-", strsplit(as.character(amu), split = "\\.")[[1]][1])) %>%
  dplyr::ungroup() %>%
  dplyr::mutate(prior_weight = pmin(1 / 100 / natural_freq, 1))

query_mass <- -1.007825 + 13.003355 + 2 * 12
estimate_chemical_formula(query_mass, probable_isotopes, penalization = "count")
#> Error in loadNamespace(x): there is no package called ‘lpSolveAPI’