create_amr_presence_absence.R

library(readr)
library(tidyr)
library(dplyr)

DF <- read_tsv("config/amrfinder.tsv", col_names = TRUE, show_col_types = FALSE)

# Replace the spaces with "_" in the column names  ------------------------

# Get the current column names
col_names <- colnames(DF)

# Replace spaces with underscores
col_names <- gsub(" ", "_", col_names)

# Assign the modified names back to the data frame
colnames(DF) <- col_names

# Remove '#' from column names
colnames(DF) <- gsub("#", "", colnames(DF))


#Remove all entries that do not contain "Neisseria gonorrhoea"
#DF <- DF %>%
#  filter(grepl("Neisseria gonorrhoeae", Scientific_name))


#change entries to "EXCLUDE" for "INTERNAL_STOP|PARTIALX|PARTIALP" in the "Method" column
#DF <- DF %>%
#  mutate(
#    Subtype = if_else(grepl("INTERNAL_STOP|PARTIALX|PARTIALP", Method), "EXCLUDE", Subtype),
#    Subclass = if_else(grepl("INTERNAL_STOP|PARTIALX|PARTIALP", Method), "EXCLUDE", Subclass)
#  )

# Filter out rows with the specified terms in the Method column
DF <- DF %>%
  filter(!grepl("INTERNAL_STOP|PARTIALX|PARTIALP", Method))


# Restructure the DF and assign 1 for present and 0 for absent  -----------

new_DF <- DF %>%
  select(BioSample, Element_symbol) %>%
  distinct() %>%
  mutate(present = 1) %>%
  pivot_wider(names_from = Element_symbol, values_from = present, values_fill = list(present = 0))

write.csv(new_DF, file = "data/biosample_gene.csv", row.names = FALSE)



# Code to check to make sure this working as expected  ----------------

count_ones_zeros <- function(df, sample_id) {
  # Filter the dataframe for the specific BioSample
  sample_row <- df[df$BioSample == sample_id, -which(names(df) == "BioSample")]
  
  # Return counts of ones and zeros
  ones_count <- sum(sample_row == 1, na.rm = TRUE)
  zeros_count <- sum(sample_row == 0, na.rm = TRUE)
  
  return(list(ones = ones_count, zeros = zeros_count))
}

# Apply the function to a specific BioSample to check for accuracy 
result <- count_ones_zeros(new_DF, "SAMD00099414")
print(result)


#check the number of unique biosamples
unique_terms <- DF %>%
  distinct(BioSample)

#check the number of unique biosamples
unique_terms <- DF %>%
  distinct(Element_symbol)