lowQuality user manual

Obtains statistics from the reads quality of a fastq file. The binary output is then read by an Rmd script that generates a html visualization of the data.

Installation

• Requires: gcc, R, Rpackages::pheatmap,rmarkdown.
• Clone the repository.
• Run make.

Running the program

Usage C executable (in folder bin):

Usage: lowQuality -i <INPUT_FILE.fq> -l <READ_LENGTH> -q [Q_THRESHOLD] -t [NUMBER_OF_TILES] -o <OUTPUT_FILE> 
Reads in a fq file (gz, bz2, z formats also accepted) and performs a set of 
quality tests. 
Output in binary format.
 -h prints help dialog.
 -i Input file [*fq|*fq.gz|*fq.bz2]. Required parameter.
 -l Read length. Length of the reads. Required parameter.
 -t Number of tiles. Optional parameter (default 96). 
 -q Minimum quality allowed. Optional parameter (default 27).
 -o Output file. Required parameter. With  *bin extension.
 -f If present, the file will be treated as a filtered file (output from filter_trim).

Usage Rmd scripts (in folder R):

1. Pro sample,

    Rscript -e "rmarkdown::render('PATH/TO/quality_report.Rmd',
         params=list(inputfile='PATH/TO/test.out'),
         output_file='PATH/TO/HTML_OUTPUT_FILE.html')"
   

2. Summary for all samples,

    Rscript -e "rmarkdown::render('PATH/TO/summary_report.Rmd', 
                params=list(inputfolder='PATH/TO/BINARY/FILES'), 
                output_file='PATH/TO/HTML_OUTPUT_FILE.html')"
   

Warning: summary_report.Rmd requires the C executable output to have the extension .bin

Depends on: rmarkdown, pheatmap R packages.

Can be used with filtered fastq files from trim_filtered

Output description

• Binary output:

  • int (4B) : read length (read_len),
  • int (4B) : number of tiles (ntiles),
  • int (4B) : minimum quality accepted (minQ),
  • int (4B) : number of possible qualities (nQ),
  • int (4B) : number of reads (nreads),
  • int (4B) : number of reads containing N's (reads_wN),
  • ntiles*int (4xntilesB) : tile tags (tile_tags),
  • nQxint (4xnQB) : quality tags (quality tags),
  • 5 x ntiles x (long int) (5xntilesx8B) : # (A,C,G,T,N) per tile with low quality (lowQ_ACGT_tile),
  • 5 x ntiles x (long int) (5xntilesx8B) : # (A,C,G,T,N) per tile (ACGT_tile),
  • (read_len+1) x (long int) ((read_len+1)x8B) : number of reads with at least m low quality nucleotides (reads_MlowQ),
  • ntiles x read_len x nQ x (long int) (ntilesxread_lenxnQx8B) : nucleotides per tile per position with a given quality (QposTile_table).
  • 5 x read_len x (long int) (5xread_lenx8B ): # (A,C,G,T,N) per position, (ACGT_pos)

• html output:

  • Table with general information ,
  • Plot 1: per base sequence quality,
  • Plot 2: # reads with at least m low Q nucleotides,
  • Plot 3: Low Q nucleotide proportion,
  • Plot 4: Average quality per position per tile,
  • Plot 5: Low Q nucleotides proportion per position per tile,
  • Plot 6: Low Q nucleotides proportion per position for all tiles.
  • Plot 7: Nucleotide content per position

• html summary:

  • Table with: # reads , # tiles, % lowQ reads, % reads with N’s for all samples.
  • Heatmap showing the average quality per position for all samples.

Example

An example is given in the folder examples. To run an example, type,

    ./bin/lowQuality -i ./examples/test.fq.bz2 -l 51 -o ./example/my_output.bin
    Rscript -e "rmarkdown::render('./R/quality_report.Rmd',
                 params=list(inputfile='./examples/my_output.bin'),
                 output_file='./examples/my_test_quality.html')"

Contributors

Paula Pérez Rubio

License

GPL v3 (see LICENSE.txt)