R/read_phosphosites.R

In casblaauw/phosphocie: Create Continuous Phosphosite Colour Representations

#' Read in PhosphoSitePlus phosphorylation site data
#'
#' Wrapper around [readr::read_tsv()] that cleans up the default phosphorylation
#' site data to have nicer columns, cleaner values, and a prepared unique site ID.
#'
#' @param path The path to the unzipped kinase substrate file.
#' @export
#'
#' @examples
#' phosphosite_path <- system.file('extdata', 'phosphorylation_site_dataset_head', package = 'phosphocie')
#' phosphosite <- phosphocie:::read_phosphosite(phosphosite_path)
#'
read_phosphosite <- function(path) {
  phosphosite_colnames <-
    c('gene',
      'protein',
      'acc_id',
      'human_chr_loc',
      'ptm_info',
      'site_grp_id',
      'organism',
      'mw_kd',
      'domain',
      'fragment_15',
      'lt_lit',
      'ms_lit',
      'ms_cst',
      'cst_cat'
    )
    # Read in data
  readr::read_tsv(path, skip = 4, col_names = phosphosite_colnames, show_col_types = FALSE) %>%
    # Drop useless columns
    dplyr::select(-c('lt_lit', 'ms_lit', 'ms_cst', 'cst_cat')) %>%
    # Separate ptm column into residue, position, and type info
    tidyr::separate(ptm_info,
                    into = c('ptm_residue', 'ptm_type'),
                    sep = '-',
                    remove = TRUE) %>%
    tidyr::separate(ptm_residue,
                    into = c('residue', 'position'),
                    sep = 1,
                    remove = FALSE,
                    convert = TRUE
    ) %>%
    # Make protein names output-safe and generate unique site ID for FASTA
    dplyr::mutate(protein = stringr::str_replace_all(protein, ' ', '_'),
                  unique_id = paste(acc_id, gene, protein, ptm_residue, sep = '|')
    ) %>%
    # Reorder columns into consistent order
    dplyr::relocate(
      dplyr::all_of(c(
        'unique_id',
        'gene',
        'protein',
        'acc_id',
        'residue',
        'position',
        'fragment_15',
        'ptm_type',
        'organism',
        'human_chr_loc',
        'ptm_residue',
        'mw_kd',
        'domain'
      )
    )) %>%
    filter_phosphosites()
}

#' Read in PhosphoSitePlus kinase substrate data
#'
#' Wrapper around [readr::read_tsv()] that cleans up the default kinase-substrate
#' data to have nicer columns, cleaner values, and a prepared unique site ID.
#'
#' @param path The path to the unzipped kinase substrate file.
#' @export
#'
#'@examples
#' kinsub_path <- system.file('extdata', 'kinase_substrate_dataset_head', package = 'phosphocie')
#' kinsub <- phosphocie:::read_kinsub(kinsub_path)
#'
read_kinsub <- function(path) {
  kinsub_colnames <-
    c('kin_gene',
      'kin_prot',
      'kin_acc_id',
      'kin_organism',
      'substrate',
      'gene_id',
      'acc_id',
      'gene',
      'organism',
      'ptm_residue',
      'site_grp_id',
      'fragment_15',
      'domain',
      'in_vivo',
      'in_vitro',
      'cst_cat'
    )

    # Read in data
  readr::read_tsv(path, skip = 4, col_names = kinsub_colnames, show_col_types = FALSE) %>%
    # Drop useless columns
    dplyr::select(-c('cst_cat')) %>%
    # Separate ptm column into residue and position columns
    tidyr::separate(ptm_residue,
                    into = c('residue', 'position'),
                    sep = 1,
                    remove = FALSE,
                    convert = TRUE
    ) %>%
    # Turn 'X'/NA columns into boolean
    dplyr::mutate(
      in_vivo = dplyr::if_else(in_vivo == 'X', TRUE, FALSE) %>% tidyr::replace_na(FALSE),
      in_vitro = dplyr::if_else(in_vitro == 'X', TRUE, FALSE) %>% tidyr::replace_na(FALSE)
    ) %>%
    # Make protein names output-safe and generate unique site ID for FASTA
    dplyr::mutate(
      substrate = stringr::str_replace_all(substrate, ' ', '_'),
      unique_id = paste(acc_id, gene, substrate, ptm_residue, sep = '|')
    ) %>%
    dplyr::relocate(
      dplyr::all_of(c(
        'unique_id',
        'gene',
        'substrate',
        'acc_id',
        'residue',
        'position',
        'fragment_15',
        'kin_gene',
        'kin_prot',
        'kin_acc_id',
        'ptm_residue',
        'organism',
        'kin_organism',
        'site_grp_id',
        'gene_id',
        'domain',
        'in_vivo',
        'in_vitro'
      )
    )) %>%
    filter_phosphosites() %>%
    filter(kin_organism == 'human')
}

#' Filter phosphosite data to human, uniprot-conformant data
#'
#' Filters out rows from non-human organisms or with non-uniprot-conformant accession ids.
#'
#' @param data Dataset from [read_phosphosite()] or [read_kinsub()]
#' @param organism_col Optional: name of data column containing organism data.
#' @param acc_id_col Optional: name of data column containing accession id data.
#'
#' @return The filtered data frame.
#' @keywords internal
#'
filter_phosphosites <- function(data, organism_col = 'organism', acc_id_col = 'acc_id') {
  uniprot_regex <- '^[OPQ][0-9][A-Z0-9]{3}[0-9]?-?\\d{1,3}$|^[A-NR-Z][0-9]([A-Z][A-Z0-9]{2}[0-9]){1,2}$'
  data %>%
    dplyr::filter(tolower(.data[[organism_col]]) == 'human') %>%
    dplyr::filter(stringr::str_detect(.data[[acc_id_col]], uniprot_regex))
}