Skip to content

Automatic Filtering, Trimming, Error Removing and Quality Control for fastq data

License

Notifications You must be signed in to change notification settings

OpenGene/AfterQC

Repository files navigation

install with bioconda

AfterQC

Automatic Filtering, Trimming, Error Removing and Quality Control for fastq data
AfterQC can simply go through all fastq files in a folder and then output three folders: good, bad and QC folders, which contains good reads, bad reads and the QC results of each fastq file/pair.
Currently it supports processing data from HiSeq 2000/2500/3000/4000, Nextseq 500/550, MiniSeq...and other Illumina 1.8 or newer formats

The author has reimplemented this tool in C++ with multithreading support to make it much faster. The new tool is called fastp and can be found at: https://github.com/OpenGene/fastp . If you prefer a C++ based tool, please use fastp instead.

An Example of Report

The report of AfterQC is a single HTML page with figures contained in. See an example: http://opengene.org/AfterQC/report.html

Features:

AfterQC does following tasks automatically:

  • Filters reads with too low quality, too short length or too many N
  • Filters reads with abnormal PolyA/PolyT/PolyC/PolyG sequences
  • Does per-base quality control and plots the figures
  • Trims reads at front and tail, according to QC results
  • For pair-end sequencing data, AfterQC automatically corrects low quality wrong bases in overlapped area of read1/read2
  • Detects and eliminates bubble artifact caused by sequencer due to fluid dynamics issues
  • Single molecule barcode sequencing support: if all reads have a single molecule barcode (see duplex sequencing), AfterQC shifts the barcodes from the reads to the fastq query names
  • Support both single-end sequencing and pair-end sequencing data
  • Automatic adapter cutting for pair-end sequencing data
  • Sequencing error estimation, and error distribution profiling

Get AfterQC

PyPy suggestion:

AfterQC is compitable with PyPy. Using PyPy to run AfterQC is strongly suggested since it can make AfterQC 3X faster than native Python (CPython).  To run with pypy, just replace python with pypy in the commands.

Simple usage:

  • Prepare your fastq files in a folder
  • For single-end sequencing, the filenames in the folder should be *R1*, otherwise you should specify --read1_flag
  • For pair-end sequencing, the filenames in the folder should be *R1* and *R2*, otherwise you should specify --read1_flag and --read2_flag
cd /path/to/fastq/folder
python path/to/AfterQC/after.py
  • three folders will be automatically generated, a folder good stores the good reads, a folder bad stores the bad reads and a folder QC stores the report of quality control
  • AfterQC will print some statistical information after it is done, such how many good reads, how many bad reads, and how many reads are corrected.
  • if you want to run AfterQC only with a single file/pair:
# with a single file
python after.py -1 R1.fq

# with a single pair
python after.py -1 R1.fq -2 R2.fq

Quality Control only

If you only want to get quality control statistics, run:

python after.py --qc_only

Gzip output

  • If the input FastQ files are gzipped, then the output will be also gzipped.  
  • If the input FastQ files are not gzipped, you can enable --gzip or -z option to force gzip compression.
  • Use --compression to change the compression level (0~9), default is 2. The better the compression, the lower the speed.

Full options:

Common options

  --version             show program's version number and exit
  -h, --help            show this help message and exit

File (name) options


  -1 READ1_FILE, --read1_file=READ1_FILE
                        file name of read1, required. If input_dir is
                        specified, then this arg is ignored.
  -2 READ2_FILE, --read2_file=READ2_FILE
                        file name of read2, if paired. If input_dir is
                        specified, then this arg is ignored.
  -7 INDEX1_FILE, --index1_file=INDEX1_FILE
                        file name of 7' index. If input_dir is specified, then
                        this arg is ignored.
  -5 INDEX2_FILE, --index2_file=INDEX2_FILE
                        file name of 5' index. If input_dir is specified, then
                        this arg is ignored.
  -d INPUT_DIR, --input_dir=INPUT_DIR
                        the input dir to process automatically. If read1_file
                        are input_dir are not specified, then current dir (.)
                        is specified to input_dir
  -g GOOD_OUTPUT_FOLDER, --good_output_folder=GOOD_OUTPUT_FOLDER
                        the folder to store good reads, by default it is the
                        same folder contains read1
  -b BAD_OUTPUT_FOLDER, --bad_output_folder=BAD_OUTPUT_FOLDER
                        the folder to store bad reads, by default it is same
                        as good_output_folder
  --read1_flag=READ1_FLAG
                        specify the name flag of read1, default is R1, which
                        means a file with name *R1* is read1 file
  --read2_flag=READ2_FLAG
                        specify the name flag of read2, default is R2, which
                        means a file with name *R2* is read2 file
  --index1_flag=INDEX1_FLAG
                        specify the name flag of index1, default is I1,
                        which means a file with name *I1* is index2 file
  --index2_flag=INDEX2_FLAG
                        specify the name flag of index2, default is I2,
                        which means a file with name *I2* is index2 file

Filter options

  -f TRIM_FRONT, --trim_front=TRIM_FRONT
                        number of bases to be trimmed in the head of read. -1
                        means auto detect
  -t TRIM_TAIL, --trim_tail=TRIM_TAIL
                        number of bases to be trimmed in the tail of read. -1
                        means auto detect
  --trim_pair_same=TRIM_PAIR_SAME
                        use same trimming configuration for read1 and read2 to
                        keep their sequence length identical, default is true
                        lots of dedup algorithms require this feature
  -q QUALIFIED_QUALITY_PHRED, --qualified_quality_phred=QUALIFIED_QUALITY_PHRED
                        the quality value that a base is qualifyed. Default 20
                        means base quality >=Q20 is qualified.
  -u UNQUALIFIED_BASE_LIMIT, --unqualified_base_limit=UNQUALIFIED_BASE_LIMIT
                        if exists more than unqualified_base_limit bases that
                        quality is lower than qualified quality, then this
                        read/pair is bad. Default 0 means do not filter reads
                        by low quality base count
  -p POLY_SIZE_LIMIT, --poly_size_limit=POLY_SIZE_LIMIT
                        if exists one polyX(polyG means GGGGGGGGG...), and its
                        length is >= poly_size_limit, then this read/pair is
                        bad. Default is 35
  -a ALLOW_MISMATCH_IN_POLY, --allow_mismatch_in_poly=ALLOW_MISMATCH_IN_POLY
                        the count of allowed mismatches when evaluating
                        poly_X. Default 5 means disallow any mismatches
  -n N_BASE_LIMIT, --n_base_limit=N_BASE_LIMIT
                        if exists more than maxn bases have N, then this
                        read/pair is bad. Default is 5
  -s SEQ_LEN_REQ, --seq_len_req=SEQ_LEN_REQ
                        if the trimmed read is shorter than seq_len_req, then
                        this read/pair is bad. Default is 35

Debubble options (not suggested for regular tasks)  
If you want to eliminate bubble artifact, turn debubble option on (this is slow, usually you don't need to do this):

  --debubble            enable debubble algorithm to remove the
                        reads in the bubbles. Default is False
  --debubble_dir=DEBUBBLE_DIR
                        specify the folder to store output of debubble
                        algorithm, default is debubble
  --draw=DRAW           specify whether draw the pictures or not, when use
                        debubble or QC. Default is on

Barcoded sequencing options

  --barcode=BARCODE     specify whether deal with barcode sequencing files, default is on
  --barcode_length=BARCODE_LENGTH
                        specify the designed length of barcode
  --barcode_flag=BARCODE_FLAG
                        specify the name flag of a barcoded file, default is
                        barcode, which means a file with name *barcode* is a
                        barcoded file
  --barcode=BARCODE     specify whether deal with barcode sequencing files,
                        default is on, which means all files with barcode_flag
                        in filename will be treated as barcode sequencing
                        files

QC options

  --qc_only             enable this option, only QC result will be output, this
                        can be much faster
  --qc_sample=QC_SAMPLE
                        sample up to qc_sample when do QC, default is 1000,000
  --qc_kmer=QC_KMER     specify the kmer length for KMER statistics for QC,
                        default is 8

Understand the report

  • AfterQC will generate a QC folder, which contains lots of figures.
  • For pair-end sequencing data, both read1 and read2 figures will be in the same folder with the folder name of read1's filename. R1 means read1, R2 means read2.
  • For single-end sequencing data, it will still have R1.
  • prefilter means before filtering, postfilter means after filtering
  • For pair-end sequencing data, After will do an overlap analysis. read1 and read2 will be overlapped when read1_length + read2_length > DNA_template_length.

Cite AfterQC

Shifu Chen, Tanxiao Huang, Yanqing Zhou, Yue Han, Mingyan Xu and Jia Gu. AfterQC: automatic filtering, trimming, error removing and quality control for fastq data. BMC Bioinformatics 2017 18(Suppl 3):80 https://doi.org/10.1186/s12859-017-1469-3