Skip to content

QuantNado

logo

QuantNado converts BAM-, bedGraph-, and VCF-derived assays into Zarr-backed stores and gives you one analysis API for region queries, feature reduction, counting, normalisation, PCA, plotting, and peak calling.

Current Workflow

  1. Run quantnado dataset create once per sample to build per-sample .zarr stores.
  2. Optionally run quantnado dataset combine to merge those stores into a single multi-sample .zarr.
  3. Open the dataset with QuantNado.open(...) or QuantNadoDataset(...).

Supported Assays

Assay Inputs Typical array keys
ATAC, ChIP, CUT&TAG BAM atac, chip_<ip>, cat_<ip>, coverage
MCC BAM viewpoint-specific arrays
METH BAM + methylation bedGraph methyl_pct, n_methylated, n_total, coverage
RNA BAM rna_fwd, rna_rev, coverage
SNP VCF.gz GT, AF, DP, MQ, coverage

Minimal Example

quantnado dataset create \
  --sample Sample-1 \
  --assay ATAC \
  --bamfile /data/ATAC-Sample-1.bam \
  --output-dir dataset \
  --chromsizes hg38.chrom.sizes

quantnado dataset create \
  --sample Sample-1_H3K27ac \
  --assay ChIP \
  --bamfile /data/Sample-1_H3K27ac.bam \
  --ip H3K27ac \
  --output-dir dataset

quantnado dataset combine \
  --stores dataset/Sample-1.zarr dataset/Sample-1_H3K27ac.zarr \
  --output dataset/combined.zarr

from quantnado import QuantNado

qn = QuantNado.open("dataset/combined.zarr")

region = qn.sel("chr21", 36_000_000, 36_010_000)
signal = qn.reduce("promoters.bed", reduction="mean", modality="coverage")
matrix, features = qn.quantify_signal("genes.gtf", feature_type="gene", assay="RNA", modality="coverage")
qn.group_by(
    ip="ip",
    treatment={"control": ["control"], "treated": ["treated"]},
    match="contains",
)
qn.info

Key Concepts

  • A dataset can be either a directory of per-sample .zarr stores or a combined .zarr.
  • assay filters samples by biological type such as RNA or ATAC.
  • modality selects a concrete array key such as coverage, rna_fwd, or chip_h3k27ac.
  • group_by() caches reusable sample-group namespaces, and subset() can intersect assay, sample, IP, and group filters.
  • Most analysis methods return xarray objects, so they stay lazy until you compute or plot.
  • reduce() returns an xr.Dataset, extract() returns an xr.DataArray, quantify_signal() returns a pandas feature matrix plus metadata from stored signal, and count_features() reserves room for backend-specific counting semantics.

Documentation