Skip to content

← Back to main page

Genome Setup

After successful initialization of SeqNado (Initialisation), use seqnado genomes to manage reference genomes.

Subcommands

Subcommand Description
list Show available genome configurations
edit Open the user genome config in $EDITOR
build Download a genome from UCSC and build indices via Snakemake
fastqscreen Generate a FastqScreen configuration file

Building a Genome

seqnado genomes build downloads the FASTA, GTF, chromosome sizes, and blacklist from UCSC, then builds Bowtie2 and STAR indices.

⚠️ Dependencies

The genome build workflow requires samtools, Bowtie2, and STAR. These are not installed by default in a basic SeqNado environment.

Use --preset to ensure dependencies are available:

  • --preset ls (recommended) — Uses Apptainer/Singularity to provide all tools
  • --preset ss — On SLURM clusters with Apptainer/Singularity
  • --preset lc — Local execution with Conda + Apptainer
  • --preset le (default) — Requires manual installation of dependencies

All examples below include --preset ls. If using a different environment, adjust accordingly.

Single genome

seqnado genomes build --name hg38 --outdir /path/to/genomes --preset ls

Multiple genomes

Comma-separate names to build several genomes in one run:

seqnado genomes build --name hg38,mm39,dm6 --outdir /path/to/genomes --preset ls

Spike-in (composite) genome

Combine a primary genome with a spike-in. This downloads both genomes, concatenates their FASTA and GTF, and builds composite indices:

seqnado genomes build --name hg38 --spikein dm6 --outdir /path/to/genomes --preset ls

The composite genome is named hg38_dm6 and spike-in chromosomes are prefixed (e.g. dm6_chr2L).

Dry run

Preview planned jobs without executing them:

seqnado genomes build --name hg38 --outdir /path/to/genomes --preset ls --dry-run

What gets built

For each genome, the workflow produces the following output structure:

<outdir>/<genome>/
├── sequence/
│   ├── <genome>.fa          # FASTA (downloaded from UCSC)
│   ├── <genome>.fa.fai      # samtools faidx index
│   └── <genome>.chrom.sizes # chromosome sizes
├── genes/
│   └── <genome>.ncbiRefSeq.gtf  # gene annotations
├── bt2_index/
│   └── <genome>.*.bt2      # Bowtie2 index
├── STAR_2.7.10b/            # STAR index directory
└── <genome>-blacklist.bed.gz  # ENCODE blacklist regions

Genome config auto-update

On successful completion, the build automatically registers the genome in ~/.config/seqnado/genome_config.json. This makes it immediately available for pipeline runs — no manual editing needed.

Build options

Option Default Description
--name, -n (required) Genome name(s), comma-separated
--outdir, -o ./genome_build Output directory
--spikein, -sp — Spike-in genome name for composite builds
--preset le Snakemake profile preset (see below)
--profile — Path to a Snakemake profile directory (overrides --preset)
--cores, -c 4 Number of cores available to Snakemake; controls max parallelism and threads per rule
--scale-resources 1.0 Scale memory/time requests
--dry-run off Preview the Snakemake DAG without executing
--verbose, -v off Print the full Snakemake command

Presets

Presets select a Snakemake execution profile. For genome builds, use a preset that includes samtools, Bowtie2, and STAR:

Preset Tools included? Environment Description
le ❌ No Local Requires manual tool installation (not recommended for genome builds)
ls ✅ Yes Local Singularity/Apptainer containers (recommended)
lc ✅ Yes Local Conda + Singularity
ld ✅ Yes Local Docker + Conda
ss ✅ Yes SLURM cluster Singularity/Apptainer containers on HPC
t ✅ Yes Local Testing/development

Listing Genomes

Show all configured genomes and their paths:

seqnado genomes list

Editing Genome Config

Open ~/.config/seqnado/genome_config.json in your editor:

seqnado genomes edit

Set $EDITOR to your preferred editor (defaults to nano).

FastqScreen Configuration

FastQ Screen checks for sample contamination by aligning reads against a set of reference genomes. SeqNado can auto-generate the fastq_screen.conf configuration file from your built genomes.

Key distinction: - --contaminant-path = input (where your contaminant reference indices are located) - --screen = output (where to save the generated config file)

Generating the config

seqnado genomes fastqscreen

This reads your genome config (~/.config/seqnado/genome_config.json), finds each genome's Bowtie2 index, and writes a fastq_screen.conf with a DATABASE entry per genome. Organism names (Human, Mouse, Drosophila, etc.) are inferred automatically from the genome prefix (e.g. hg38 → Human, mm39 → Mouse).

Adding contaminant databases

By default, the command prompts for a path to contaminant reference files. If provided, it adds common contaminant screens (E. coli, PhiX, rRNA, adapters, etc.).

A pre-built set of contaminant references is available. Download and extract it:

wget https://userweb.molbiol.ox.ac.uk/public/project/milne_group/seqnado/genomes/fastqscreen_reference.tar.gz
tar -xzf fastqscreen_reference.tar.gz

Then generate the config with contaminants:

# Quick example: contaminants in current directory
seqnado genomes fastqscreen --contaminant-path ./fastqscreen_reference

# Specify custom output path
seqnado genomes fastqscreen --contaminant-path ./fastqscreen_reference --screen /path/to/my_fastq_screen.conf

# Skip contaminants entirely
seqnado genomes fastqscreen --no-contaminants

The contaminant directory should contain Bowtie2 indices organised in subdirectories:

fastqscreen_reference/
├── E_coli/Ecoli.*bt2
├── PhiX/phi_plus_SNPs.*bt2
├── rRNA/GRCm38_rRNA.*bt2
├── Vectors/Vectors.*bt2
├── Adapters/Contaminants.*bt2
└── ...

Options

Option Default Description
--screen, -s ~/.config/seqnado/fastq_screen.conf Output path: Where to write the generated fastq_screen.conf
--contaminant-path (prompted) Input path: Directory containing contaminant Bowtie2 indices
--threads, -t 8 Bowtie2 threads used by FastQ Screen
--no-contaminants off Skip contaminant databases

See Also: