purrPore 0.1.0

Merge ONT nanopore data from POD5 and FASTQ for per-pore QC

purrPore

Merge Oxford Nanopore (ONT) sequencing data from POD5 and FASTQ for per-pore QC.

purrPore loads per-read metadata from POD5 files (channel, well, duration, sample rate) and merges it with FASTQ-derived metrics (length, mean Q, GC). It aggregates by (channel, well) for true per-pore QC—each channel has up to 4 wells (nanopores). Includes N50 per pore, pore dropout detection, and flow cell heatmaps (physical MinION layout).

Install

pip install -e .

Requires Python ≥3.8.

CLI

purrPore run reads.fastq pod5_dir/ -o qc_out/
purrPore run reads.fastq pod5_dir/ -o qc_out/ --heatmap   # also generate heatmaps

Python API

import purrPore

results = purrPore.run_qc("reads.fastq", "pod5_dir/")

# Merged read-level table (FASTQ + POD5 metadata)
print(results.merged)

# Per-pore: reads, yield_bp, mean_len, mean_q, mean_gc, median_speed, active_time_s, n50
print(results.per_pore)

# Flowcell: total_reads, total_yield_bp, mean_q, active_pores, dropout_pores
print(results.flowcell)

# Pores (channel, well) that produced 0 reads (dropout)
print(results.dropout_pores[:20])

# Write outputs (CSV, Parquet, dropout list)
purrPore.write_qc_results(results, "qc_out/")

# Flow cell heatmap (128×16 pore grid: 4 wells per channel)
purrPore.plot_flowcell_heatmap(results.per_pore, value_col="yield_bp", out_path="heatmap.png")

Features

• Flow cell heatmap: Physical MinION layout (128×16 pore grid: 4 wells per channel)
• Read N50 per pore: Length such that 50% of pore yield is in reads of that length or longer
• Pore dropout detection: Pores (channel, well) with 0 reads (written to dropout_pores.txt)

Data

• FASTQ: read_id, read_length, mean_q, gc
• POD5: read_id, channel, well, duration_s, ...
• Per-pore: channel, well, reads, yield_bp, mean_len, mean_q, mean_gc, median_speed, active_time_s, n50
• Flowcell: total_reads, total_yield_bp, mean_q, active_pores, dropout_pores

License

MIT.