QC pipeline - a-lud/nf-pipelines GitHub Wiki
This is a quality-control (QC) sub-workflow for paired-end short-read sequence data.
The current version of the QC pipeline requires the following arguments.
--seqdir string Directory path containing paired-end short-read FASTQ files.
--platform string Specify the sequencing platform. Options: illumina, mgi.
--krakendb string Directory path to pre-installed Kraken2 database.
--bq_phred integer Quality value that a base is qualified. Default phread quality >= 15.
--n_base_limit integer Number of N's in a read before read-pair is removed. Default is 5
--average_qual integer Average quality required by a read to not be filtered out. Default 0 (no minimum avg. quality).
--length_required integer Reads shorter than this length will be discarded. Default is 15.
This argument is merely the directory path to the shor-read sequencing data. The read files are expected to be
paired-end and gzipped with the extension _R{1,2}. If the files do not match match these requirements, the
pipeline will fail to find the files and error out. Valid file names are shown below.
- sample1_R1.fastq.gz
- sample1_R2.fq.gz
- sample.information_R1.fastq.gz
- sample_information_R2.fq.gz
This argument requires the specification of the sequencing platform the data was generated on. Currently illumina and mgi are
valid options. This is important for the Fastp process, as it can correct MGI sequence headers that don't play well with downstream
software (e.g. BAM files).
Provide the path to a pre-installed Kraken2 database. The Phoenix compute nodes do not have access to the internet, so it is easier to
simply download this ahead of time and pass the path to the directory. Pre-compiled Kraken2 databases can be downloaded from this website.
This is an argument specifically for Fastp (as are all the following arguments), corresponding to the qualified_quality_phred argument.
This argument controls the quality a base needs to be to be qualified. If a base-quality is less than this value, it will be removed.
The default value is 15.
This argument controls the number of N sequences that are allowed to be in a read before it is filtered out. The default value is 5.
This argument controls the minimum average quality of a read. If the average quality of a read drops below this value, it will be filtered out.
The default value is 0, which means no filtering will take place and all reads will pass.
The final argument controls the minimum length required by a read. If a read falls below this value (default: 15), it will be filtered out.
I recommend setting this to around half the expected length.
The QC sub-workflow generates a number of output files, however the most useful is likely to be the multiqc_report.html. This is an
aggregation of all the QC results.
qc-results/
├── reports
│ ├── DAG.svg
│ ├── report.html
│ ├── timeline.html
│ └── trace.txt
└── out_prefix
├── fastp
│ ├── id.fastp.html
│ ├── id.fastp.json
│ ├── id_R1.fastq.gz
│ ├── id_R1.unpaired.gz
│ ├── id_R2.fastq.gz
│ └── id_R2.unpaired.gz
├── fastqc
│ ├── id_R1_fastqc.html
│ ├── id_R1_fastqc.zip
│ ├── id_R2_fastqc.html
│ └── id_R2_fastqc.zip
├── kraken2
│ ├── id_1.fastq.gz
│ ├── id_2.fastq.gz
│ ├── id-classified_1.fastq.gz
│ ├── id-classified_2.fastq.gz
│ └── id.report
└── multiqc
├── multiqc_data
│ ├── multiqc_citations.txt
│ ├── multiqc_data.json
│ ├── multiqc_fastp.txt
│ ├── multiqc_fastqc.txt
│ ├── multiqc_general_stats.txt
│ ├── multiqc.log
│ └── multiqc_sources.txt
└── multiqc_report.html
There will be a sub-directory for each QC-process with the relevant output files from each piece of software if you so wish to examine individual files/the speific outputs.