🌶️ sracha 🌶️

Fast SRA downloader and FASTQ converter, written in pure Rust.

Features

Parallel downloads -- chunked HTTP Range requests with multiple connections
Native VDB parsing -- pure Rust, zero C dependencies
Integrated pipeline -- download, convert, and compress in one command
Project-level accessions -- pass a BioProject (PRJNA) or study (SRP) to download all runs
Accession lists -- batch download from a file with --accession-list
Parallel gzip or zstd -- pigz-style block compression via rayon
FASTA output -- drop quality scores with --fasta
SRA and SRA-lite -- full quality or simplified quality scores
Split modes -- split-3, split-files, split-spot, interleaved
Resumable downloads -- automatically resumes interrupted transfers
File validation -- verify SRA file integrity with sracha validate

How it works

sracha get runs the full pipeline in one command:

Resolve -- looks up the accession via direct S3 URL (with SDL API fallback)
Download -- fetches the .sra file with parallel chunked HTTP Range requests
Parse -- reads the KAR archive and decodes VDB columns (READ, QUALITY, READ_LEN, NAME)
Output -- formats FASTQ (or FASTA) records and compresses with parallel gzip/zstd

Quick start

# Download, convert, and compress in one shot
sracha get SRR28588231

# Download all runs from a BioProject
sracha get PRJNA675068

# Batch download from an accession list
sracha get --accession-list SRR_Acc_List.txt

# Just download
sracha fetch SRR28588231

# Convert a local .sra file
sracha fastq SRR28588231.sra

# Show accession info
sracha info SRR28588231

# Validate a downloaded file
sracha validate SRR28588231.sra

See the Getting Started guide for more examples, or the CLI Reference for all options.

Installation

From binary releases

Download pre-built binaries from the releases page.

From source

Requires Rust 1.92+.

cargo install --git https://github.com/rnabioco/sracha-rs sracha

With Bioconda

pixi add --channel bioconda sracha