Demultiplex into FASTQ files from parquet file

Output read-1s into one FASTQ file per sample. Optionally clip off at poly(A) tail or low quality region. Read names have UMI and barcode appended.

Usage

demux_reads(
  out_dir,
  in_file,
  sample_names = NULL,
  clip = FALSE,
  clip_min_untemplated = 12,
  clip_min_length = 20,
  max_t_read_1 = 20,
  min_t = 0,
  verbose = FALSE
)

Arguments

out_dir

Directory to save fastq files.

in_file

Parquet filename, for file produced by ingest_read_pairs.

clip

Should reads be clipped. Don't use this if reads are intended for Tail Tools.

clip_min_untemplated

If clipping and a poly(A) tail plus expected suffix bases (UMI, barcode) was detected with at least this length, clip to this length. Otherwise clip at end of good quality region.

clip_min_length

If clipped read is shorter than this, discard it.

max_t_read_1

If read 1 had a poly(T) span longer than this, the read is discarded.

min_t

If read 2 had a poly(T) length less than this in read 2, the read is discarded.

Details

Various total read counts are provided in the output stats.parquet file. If clipping is disabled, statistics will reflect what would have been done if clipping was used.