PeakPicker

HPLC chromatogram peak detection, baseline correction, deconvolution, and calibration-curve quantification for Agilent ChemStation data.

pip install hplc-peakpicker

Features

• File reading: Agilent Chemstation .ch files (format 130/131), .D folder scanning, Rainbow API support
• Baseline correction: ArPLS, weighted spline, hybrid valley-based strategies with quality evaluation
• Peak detection: Two-pass detection with configurable parameters
• Peak deconvolution: Gaussian and EMG (Exponentially Modified Gaussian) fitting via lmfit
• Quantification: Calibration curves, sample parsing, statistical analysis, batch processing
• Export: Excel reports and publication-quality plots

Installation

pip install hplc-peakpicker        # from PyPI
pip install -U hplc-peakpicker     # upgrade to the latest release

From a clone (development):

conda activate PeakPicker
pip install -r requirements.txt

Update notice

Building a workflow checks PyPI at most once a day (1.5 s timeout, silent when offline) and prints a one-line upgrade hint to stderr when a newer release exists. Ask explicitly with peakpicker.check_for_update(); turn the check off with PEAKPICKER_NO_UPDATE_CHECK=1 (it is also skipped under CI and pytest).

Project Structure

src/
  peakpicker/          # Main package (clean architecture)
    application/       # Workflows and batch processing
    baseline/          # Baseline correction strategies
    config/            # Configuration dataclasses
    domain/            # Domain models and enums
    infrastructure/    # File I/O, exporters, signal processing
    peak_analysis/     # Peak detection and deconvolution
    quant/             # Quantification methods (public API: QuantMethod, StandardCurve)

Lab-specific code: plugins

The library is generic. Sample-name conventions, quantification presets and method-selection rules for a particular lab live outside this repository as plugins and method YAMLs — see docs/PLUGINS.md.

Related

• sci-toolkit — the lab's research skill bundle. Its scripts/hplc_parser.py is a stdlib-only port of this project's peak detection for quick trace reads; its lab-data-analysis skill routes full quantification here.

Dependencies

numpy, pandas, scipy, matplotlib, openpyxl, lmfit, pybaselines, rainbow-api