OCMSExplorer: OCMSlooksy

## code to prepare `example_data` dataset goes here

usethis::use_data("example_data")

# save_dir <- "/gfs/devel/syen/OCMSlooksy/data-raw"
save_dir <- "./data-raw"
# create database
con <- RSQLite::dbConnect(RSQLite::SQLite(), file.path(save_dir, "DSS_DB"))

# writing database table into list to save as RData file
# extract data tables
table_ls <- RSQLite::dbListTables(con)

example_data <- list()
for(i in 1:length(table_ls)) {
  query <- sprintf("SELECT * FROM %s", table_ls[i])
  entry <- RSQLite::dbGetQuery(con, query)

  example_data[[table_ls[i]]] <- entry
}
# close connection
RSQLite::dbDisconnect(con)

# add metadata to list
metadata <- read.csv(file.path(save_dir, "dss_metadata.tsv"), sep="\t")

# remove empty columns
metadata <- metadata[colSums(!is.na(metadata)) > 0]

# remove columns with one unique entry
nval <- apply(metadata, 2, function(x) length(unique(x)))
metadata <- metadata[which(nval > 1)]

# rename phenotype to treatment
colnames(metadata)[which(colnames(metadata) == 'Phenotype')] <- 'Treatment'

# add column to easily filter out failed samples
# that were idenitifed in post-hoc analyses
omit <- c('stool-2DSS__12',
          'stool-2DSS__25',
          'stool-2DSS__28',
          'stool-2DSS__30',
          'stool-3DSS__2',
          'stool-3DSS__7',
          'stool-3DSS__11',
          'stool-3DSS__16',
          'stool-3DSS__21',
          'stool-3DSS__22',
          'stool-3DSS__23',
          'stool-3DSS__24',
          'stool-3DSS__26',
          'stool-3DSS__29',
          'stool-4DSS__1')

metadata$seq_success <- ifelse(metadata$sampleID %in% omit, FALSE, TRUE)

# add interaction group
metadata$treat.geno <- paste(metadata$Treatment, metadata$Genotype, sep='.')
example_data$metadata <- metadata

save(example_data, 'example_data', file = './data/example_data.RData')