selectSamples.R

filter_microarray <- function(allTargets, seed = 77129830) {
  # Configurar la semilla aleatoria
  set.seed(77129830)
  
  # Filtrar las filas donde 'time' no sea 'hour 2'
  filtered <- subset(allTargets, time != "hour 2")
  
  # Dividir el dataset por grupos únicos de 'infection' + 'agent'
  filtered$group <- interaction(filtered$infection, filtered$agent)
  
  # Seleccionar 4 muestras al azar de cada grupo
  selected <- do.call(rbind, lapply(split(filtered, filtered$group), function(group_data) {
    if (nrow(group_data) > 4) {
      group_data[sample(1:nrow(group_data), 4), ]
    } else {
      group_data
    }
  }))
  
  # Obtener los índices originales como nombres de las filas seleccionadas
  original_indices <- match(selected$sample, allTargets$sample)
  
  # Modificar los rownames usando 'sample' y los índices originales
  rownames(selected) <- paste0(selected$sample, ".", original_indices)
  
  # Eliminar la columna 'group' y devolver el resultado
  selected$group <- NULL
  return(selected)
}



# Simular el dataset basado en la descripción proporcionada
allTargets <- data.frame(
  sample = c("GSM944831", "GSM944838", "GSM944845", "GSM944852", "GSM944859",
             "GSM944833", "GSM944840", "GSM944847", "GSM944854", "GSM944861",
             "GSM944834", "GSM944841", "GSM944848", "GSM944855", "GSM944862",
             "GSM944832", "GSM944839", "GSM944846", "GSM944853", "GSM944860",
             "GSM944835", "GSM944842", "GSM944849", "GSM944856", "GSM944863",
             "GSM944836", "GSM944843", "GSM944850", "GSM944857", "GSM944864",
             "GSM944837", "GSM944844", "GSM944851", "GSM944858", "GSM944865"),
  infection = c(rep("uninfected", 15), rep("S. aureus USA300", 20)),
  time = c(rep("hour 0", 15), rep("hour 2", 5), rep("hour 24", 15)),
  agent = c(rep("untreated", 5), rep("linezolid", 5), rep("vancomycin", 5),
            rep("untreated", 5), rep("untreated", 5), rep("linezolid", 5), rep("vancomycin", 5))
)

# Aplicar la función (cambiar 123 por vuestro ID de la UOC u otro número que podáis escribir en el documento)
result <- filter_microarray(allTargets, seed=77129830)


-----