a b/R/write_dummy_data.R
1
#' Write random sequences to fasta file
2
#' 
3
#' Create random sequences from predefined vocabulary and write to fasta file.
4
#' 
5
#' @param file_path Output directory; can also be a file name but only possible if \code{write_to_file_path = TRUE} and 
6
#' \code{num_files = 1}).
7
#' @param num_files Number of files to create.
8
#' @param header Fasta header name. 
9
#' @param seq_length Length of one sequence. If vector longer than 1, will randomly sample from that vector. 
10
#' @param num_seq Number of sequences per file.
11
#' @param fasta_name_start Beginning string of file name. Output files are named fasta_name_start + _i.fasta where i is an integer index.
12
#' @param write_to_file_path Whether to write output directly to \code{file_path}, i.e. file_path is not a directory.
13
#' @param prob Probability of each character in the \code{vocabulary} to be sampled. If `NULL` each character has same probability.
14
#' @param vocabulary Set of characters to sample sequences from. 
15
#' @examples
16
#' path_output <- tempfile()
17
#' dir.create(path_output)
18
#' create_dummy_data(file_path = path_output,
19
#'                   num_files = 3,
20
#'                   seq_length = 11, 
21
#'                   num_seq = 5,                   
22
#'                   vocabulary = c("a", "c", "g", "t"))
23
#' list.files(path_output)                
24
#'    
25
#' @returns None. Writes data to files.     
26
#' @export
27
create_dummy_data <- function(file_path,
28
                              num_files,
29
                              header = "header", 
30
                              seq_length, 
31
                              num_seq,
32
                              fasta_name_start = "file",
33
                              write_to_file_path = FALSE,
34
                              prob = NULL,
35
                              vocabulary = c("a", "c", "g", "t")) {
36
  
37
  if (!is.null(prob)) {
38
    stopifnot(length(prob) == length(vocabulary))
39
    stopifnot(abs(1 - sum(prob)) < 1e-10)
40
  }
41
  
42
  if (write_to_file_path) {
43
    stopifnot(num_files == 1)
44
  }
45
  
46
  if (!dir.exists(file_path) & !write_to_file_path) {
47
    dir.create(file_path)
48
  }
49
  
50
  for (i in 1:num_files){
51
    df <- data.frame(Header = NULL, Sequence = NULL)
52
    for (j in 1:num_seq) {
53
      nuc_seq <- random_seq(vocabulary = vocabulary, seq_length = seq_length, prob = prob)
54
      new_row <- data.frame(Header = header, Sequence = nuc_seq)
55
      df <- rbind(df, new_row)
56
    }
57
    df$Header <- as.character(df$Header)
58
    df$Sequence <- as.character(df$Sequence)
59
    if (write_to_file_path) {
60
      out.file <- file_path
61
    } else {
62
      file_name <- paste0(fasta_name_start, "_", i, ".fasta")
63
      out.file <- file.path(file_path, file_name)
64
    }
65
    microseq::writeFasta(fdta = dplyr::as_tibble(df), 
66
                         out.file =  out.file)
67
  }
68
}
69
70
random_seq <- function(vocabulary, seq_length, prob = NULL) {
71
  if (length(seq_length) > 1) {
72
    seq_length <- sample(seq_length, 1)
73
  }
74
  sample(vocabulary, seq_length, replace = TRUE, prob = prob) %>% paste(collapse = "")
75
}