preprocessing¶

abutils provides several functions for preprocessing sequence data, including merging paired-end FASTQ files. The preprocessing module can handle different file naming schemas (Illumina and Element) and supports various merging algorithms.

The primary function is abutils.tl.merge_fastqs(), which handles the process of organizing, grouping, and merging paired-end FASTQ files.

preprocessing method	function
Merge paired-end reads	abutils.tl.merge_fastqs()
Merge with fastp	abutils.tl.merge_fastqs_fastp()
Merge with vsearch	abutils.tl.merge_fastqs_vsearch()

examples¶

merge paired-end FASTQ files using fastp

By default, merge_fastqs() uses fastp to merge paired-end reads, which also performs quality filtering and adapter trimming.

import abutils

# merge paired-end FASTQ files from a directory
merged_files = abutils.tl.merge_fastqs(
    files='path/to/fastq/directory',
    output_directory='path/to/output',
    schema='illumina',  # file naming schema: 'illumina' or 'element'
    compress_output=True,  # output gzipped FASTQ files
    verbose=True
)

customize quality trimming and adapter removal

You can customize the quality trimming and adapter removal parameters.

import abutils

# merge with customized quality trimming and adapter removal
merged_files = abutils.tl.merge_fastqs(
    files='path/to/fastq/directory',
    output_directory='path/to/output',
    trim_adapters=True,
    adapter_file='path/to/adapters.fasta',  # custom adapter sequences
    quality_trim=True,
    window_size=5,  # sliding window size
    quality_cutoff=15,  # quality threshold
    minimum_overlap=20,  # minimum overlap between reads
    log_directory='path/to/logs'  # save fastp reports
)

direct low-level merging with specific algorithms

For more control, you can directly use the algorithm-specific merge functions. This requires specifying each of the input files and output file paths, rather than simply input and output directories.

import abutils

# merge a specific pair of files with fastp
abutils.tl.merge_fastqs_fastp(
    forward='path/to/sample_R1.fastq.gz',
    reverse='path/to/sample_R2.fastq.gz',
    merged='path/to/output/sample.fastq.gz',
    minimum_overlap=25,
    allowed_mismatches=3,
    trim_adapters=True,
    quality_trim=True
)

# or with vsearch
abutils.tl.merge_fastqs_vsearch(
    forward='path/to/sample_R1.fastq.gz',
    reverse='path/to/sample_R2.fastq.gz',
    merged_file='path/to/output/sample.fastq',
    output_format='fastq',
    minimum_overlap=25,
    allowed_mismatches=3
)

preprocessing¶

examples¶

api¶