jxshi/seqkit

Toolkit for manipulating FASTA and SAM files

Seqkit

Seqkit is a suite of software utilities for manipulating and analyzing common genome sequencing data types (FASTA, SAM). Seqkit is written in Rust, and uses rust-htslib for reading and writing BAM files. Seqkit is divided into two utilities: fasta and sam. Each utility provides various useful subcommands. For a complete listing type the command name without arguments into your shell.

Features

For FASTA/FASTQ files, Seqkit can:

• Convert between FASTA, FASTQ and raw sequence-per-line formats
• Extract sample barcodes and UMIs from multiplexed FASTQ sequencing data
• Demultiplex FASTQ sequencing data
• Trim FASTQ sequencing data by per-base quality (BASEQ) values
• Mask low quality bases in FASTQ sequencing data
• Replace FASTQ read identifiers with compact numeric IDs

For BAM files, Seqkit can:

• Extract reads from name-sorted or position-sorted BAM files
• Calculate a histogram of fragment lengths
• Calculate statistics about unaligned, aligned and duplicate-flagged reads

Installation

Install Rust (version 1.31 or later). Then run the following command:

cargo install --force --git https://github.com/annalam/seqkit

Examples

Extracting reads

Extract reads from a name-sorted or position-sorted BAM file called tumor.bam. Paired end reads are written in gzip-compressed FASTQ format into output files tumor_1.fq.gz and tumor_2.fq.gz, which are automatically created. Orphan reads are written into output file tumor.fq.gz. The second parameter specifies the prefix used for the output file names.

sam to fastq tumor.bam tumor

Demultiplex a pooled sequencing run

Extract UMIs and demultiplex Illumina sequencing data where both the sample barcode and UMI are stored in the adapter:

fasta demultiplex sample_sheet.tsv
  <(fasta add barcode multiplexed_R1.fq.gz multiplexed_I1.fq.gz)
  <(fasta add barcode multiplexed_R2.fq.gz multiplexed_I1.fq.gz)