A seqtk-style toolkit for sequence analysis. By no means feature complete. In fact largely contains features other authors have not implemented in their respective tools.

Seqhax is packaged in bioconda, please mamba install -c bioconda seqhax if conda is working for you. Otherwise, static binaries are uploaded by the CI to github releases. Or, you can build it yourself:

sudo apt-get install zlib1g-dev libhts-dev libhtscodecs-dev  # Or whatever equivalent for your system
git clone https://github.com/kdmurray91/seqhax.git
cd seqhax
mkdir build && cd build
cmake ..
make
make install

To make static binaries, one can use cmake -DSTATIC_BUILD=On .. in the above series of commands.

Any other issues, file a bug report on github.

There are two commands in this package: seqhax and htshax.

The seqhax command contains various sequence file manipulation subcommands. The commands, along with a synopsis of their actions, are displayed when one types seqhax with no arguments. At the time of writing, these were:

$ seqhax
USAGE:
    seqhax PROGRAM [options]

where PROGRAM is one of:
    anon       -- Rename sequences by a sequential number
    convert    -- Convert between FASTA and FASTQ formats
    clihist    -- Records a histogram of stdin.
    filter     -- Filter reads from a sequence file
    pairs      -- (De)interleave paired end reads
    pecheck    -- Check that paired end reads match properly
    preapp     -- Prepend or append string to sequences
    randseq    -- Generate a random sequence file
    rebarcode  -- Add index sequences from header to the start of reads
    stats      -- Basic statistics about sequence files
    trunc      -- Truncate sequences

The usage of each subcommand can be obtained using the -h flag to that subcommand, e.g. seqhax preapp -h.

Some brief description of the more involved commands is below.

Interleaves or de-interleaves paired reads. Converts between the following forms:

• Separate R1/R2 paired files and single read read file.
• "Strict" interleaved file, where failed/missing reads are replaced with a single 'N'.
• "Broken paired" interleaved files, where failed/missing reads are simply not present.

Checks that read pairs are correctly matched, between split (R1 & R2) files, or interleaved files. Optionally, can be used to join multiple R1/R2 from the sample sample into a single interleaved file, while checking read IDs match.

htshax consists of various utilities for the handling of HTS formats via HTSlib. These are mostly things that I wish samtools/bcftools would do, but don't.

$ htshax
Usage
    htshax PROGRAM [options]

where PROGRAM is one of:

bcfhist -- Calculate histograms on MAF allele freq and missingness across samples

Create histograms over various V/BCF metrics used in variant filtering (QUAL, ALLELE_FREQ, F_MISSING, INFO/DP). Nominally bcftools stats could do this, but it doesn't summarise missingness, so I wrote this quick tool to do so.

GPL v3 (see ./LICENSE).

Copyright (c) 2014-2023 Kevin Murray.