Docs
Getting Started

Getting Started

Installation

Build from source

Requires Go 1.25+ with CGO enabled (for SQLite).

git clone https://github.com/nishad/srake.git
cd srake
go build -tags "sqlite_fts5" -o srake ./cmd/srake
./srake --help

Go install 

go install -tags "sqlite_fts5" github.com/nishad/srake/cmd/srake@latest
The sqlite_fts5 build tag is required for full-text search support.

Quick Start

Ingest SRA metadata 

# Auto-select the best archive from NCBI
srake ingest --auto

# Or ingest a local file
srake ingest --file /path/to/NCBI_SRA_Metadata.tar.gz

# Ingest with filters (human Illumina RNA-Seq only)
srake ingest --auto \
  --taxon-ids 9606 \
  --platforms ILLUMINA \
  --strategies RNA-Seq

Build a search index 

# Build Bleve full-text index
srake index --build

# Optionally include vector embeddings for semantic search
srake index --build --with-embeddings

Search 

# Basic search
srake search "homo sapiens"

# Search with filters
srake search "cancer" \
  --organism "homo sapiens" \
  --platform ILLUMINA \
  --library-strategy RNA-Seq

# Semantic vector search
srake search "tumor gene expression" --search-mode vector

# Export results as JSON
srake search "RNA-Seq" --format json --output results.json

Start the API server 

srake server --port 8080

# Query the API
curl "http://localhost:8080/api/v1/search?q=cancer&limit=10"
curl "http://localhost:8080/api/v1/stats"

Data paths

SRAKE follows the XDG Base Directory Specification:

DirectoryDefault PathPurpose
Config~/.config/srake/Configuration file
Data~/.local/share/srake/Database and models
Cache~/.cache/srake/Downloads and search index
State~/.local/state/srake/Resume checkpoints

Override with environment variables:

SRAKE_DB_PATH=/fast/ssd/srake.db srake ingest --auto
SRAKE_CACHE_HOME=/tmp/srake srake index --build

See Configuration for the full list.

Next steps

CLI Reference
All commands and flags
Features
Search, filtering, export
API Reference
REST API endpoints
DocsFeatures