R/select_sirius_columns.R

In taxonomicallyinformedannotation/tima-r: Taxonomically Informed Metabolite Annotation

#' @title Select sirius columns (canopus)
#'
#' @description This function selects sirius columns (canopus)
#'
#' @param df Dataframe
#' @param sirius_version Sirius version
#'
#' @return NULL
#'
#' @export
#'
#' @examples NULL
select_sirius_columns_canopus <- function(df, sirius_version) {
  df <- df |>
    tidytable::mutate(feature_id = switch(sirius_version,
      "5" = harmonize_names_sirius(id),
      "6" = mappingFeatureId
    )) |>
    tidytable::select(tidytable::any_of(
      c(
        "feature_id",
        "candidate_adduct" = "adduct",
        "candidate_structure_molecular_formula" = "molecularFormula",
        "feature_pred_tax_npc_01pat_val" = "NPC#pathway",
        "feature_pred_tax_npc_01pat_score" = "NPC#pathway Probability",
        "feature_pred_tax_npc_02sup_val" = "NPC#superclass",
        "feature_pred_tax_npc_02sup_score" = "NPC#superclass Probability",
        "feature_pred_tax_npc_03cla_val" = "NPC#class",
        "feature_pred_tax_npc_03cla_score" = "NPC#class Probability",
        "feature_pred_tax_cla_01kin_val" = "ClassyFire#TODO",
        "feature_pred_tax_cla_01kin_score" = "ClassyFire#TODO Probability",
        "feature_pred_tax_cla_02sup_val" = "ClassyFire#superclass",
        "feature_pred_tax_cla_02sup_score" =
          "ClassyFire#superclass probability",
        "feature_pred_tax_cla_03cla_val" = "ClassyFire#class",
        "feature_pred_tax_cla_03cla_score" = "ClassyFire#class Probability",
        "feature_pred_tax_cla_04dirpar_val" =
          "ClassyFire#most specific class",
        "feature_pred_tax_cla_04dirpar_score" =
          "ClassyFire#most specific class Probability"
      )
    ))
  return(df)
}

#' @title Select sirius columns (formulas)
#'
#' @description This function selects sirius columns (formulas)
#'
#' @include harmonize_names_sirius.R
#'
#' @param df Dataframe
#' @param sirius_version Sirius version
#'
#' @return NULL
#'
#' @export
#'
#' @examples NULL
select_sirius_columns_formulas <- function(df, sirius_version) {
  df <- df |>
    tidytable::mutate(feature_id = switch(sirius_version,
      "5" = harmonize_names_sirius(id),
      "6" = mappingFeatureId
    )) |>
    tidytable::mutate(
      candidate_structure_exact_mass = as.numeric(ionMass) -
        as.numeric(`massErrorPrecursor(ppm)`) *
          as.numeric(ionMass) *
          1E-6,
      candidate_structure_error_mz = as.numeric(ionMass) *
        as.numeric(`massErrorPrecursor(ppm)`) *
        1E-6
    ) |>
    tidytable::select(tidytable::any_of(c(
      "feature_id",
      "candidate_adduct" = "adduct",
      "candidate_structure_molecular_formula" = "molecularFormula",
      "candidate_structure_exact_mass",
      "candidate_structure_error_mz",
      "candidate_score_sirius_zodiac" = "ZodiacScore",
      "candidate_score_sirius_sirius" = "SiriusScore",
      "candidate_score_sirius_tree" = "TreeScore",
      "candidate_score_sirius_isotope" = "IsotopeScore",
      "candidate_score_sirius_intensity" = "explainedIntensity",
      "candidate_count_sirius_peaks_explained" = "numExplainedPeaks"
    )))
  return(df)
}

#' @title Select sirius columns (structures)
#'
#' @description This function selects sirius columns (structures)
#'
#' @param df Dataframe
#' @param sirius_version Sirius version
#'
#' @return NULL
#'
#' @export
#'
#' @examples NULL
select_sirius_columns_structures <- function(df, sirius_version) {
  df <- df |>
    tidytable::select(tidytable::any_of(c(
      "feature_id",
      "candidate_adduct" = "adduct",
      "candidate_structure_name" = "name",
      "candidate_structure_smiles_no_stereo" = "smiles",
      "candidate_structure_inchikey_no_stereo" = "InChIkey2D",
      "candidate_structure_molecular_formula" = "molecularFormula",
      "candidate_structure_xlogp" = "xlogp",
      # ISSUE see #147
      "candidate_score_sirius_confidence" = switch(sirius_version,
        "5" = "ConfidenceScore",
        "6" = "ConfidenceScoreApproximate"
      ),
      "candidate_score_sirius_csi" = "CSI:FingerIDScore",
      "candidate_score_sirius_msnovelist" = "ModelScore"
    ))) |>
    tidytable::distinct() |>
    tidytable::mutate(
      candidate_library = "SIRIUS",
      candidate_structure_tax_npc_01pat = NA_character_,
      candidate_structure_tax_npc_02sup = NA_character_,
      candidate_structure_tax_npc_03cla = NA_character_,
      candidate_structure_tax_cla_chemontid = NA_character_,
      candidate_structure_tax_cla_01kin = NA_character_,
      candidate_structure_tax_cla_02sup = NA_character_,
      candidate_structure_tax_cla_03cla = NA_character_,
      candidate_structure_tax_cla_04dirpar = NA_character_
    )
  return(df)
}