Metabodecon-Rust

Metabodecon is a work-in-progress project that provides functionality for handling and processing 1D NMR spectra. A more limited version of this was already implemented in R here. This project aims to improve on that by providing a more efficient and flexible implementation in Rust as well as additional features that are not present in the R version.

Features

Current planned and implemented features include (additional suggestions are welcome as issues):

• Efficient representation of 1D NMR spectra
• Read 1D NMR spectra from various formats
  • Internal HDF5 format
  • Bruker
  • JCAMP-DX
• Storing 1D NMR spectra in HDF5 files which are easier to handle and share than the raw data formats
• Preprocessing of 1D NMR spectra
  • Removal of the solvent signal
  • Smoothing of the signal intensities (various methods)
• Peak detection in 1D NMR spectra
  • Curvature analysis
• Fitting of peaks to deconvolute 1D NMR spectra
  • Lorentzian functions
    • Analytical solution to the system of equations for the Lorentzian parameters using the detected peaks
• Alignment of 1D NMR spectra using the deconvoluted signals
• Python bindings
  • Basic bindings
  • Complete Python package
• R bindings
  • Basic bindings
  • Complete R package
• Command line interface for the core crate
• Graphical user interface for the core crate

Getting Started

Installation

For now, the project is in a very early stage. However, you can already include the library crate in your own project by adding the following to your Cargo.toml:

[dependencies]
metabodecon = { git = "https://github.com/SombkeMaximilian/metabodecon-rust" }

To install the Python bindings, follow these steps:

• activate your virtual environment

• install the maturin package using pip install maturin

• navigate to the metabodecon-python crate and run the following command:

  maturin develop --release

Examples

Here is a simple example of how to use the library in Rust:

use metabodecon::{deconvolution, spectrum};

fn main() {
    // Read a spectrum from Bruker TopSpin format
    let mut spectrum = spectrum::BrukerReader.read_spectrum(
        "path/to/spectrum",
        // Experiment Number
        10,
        // Processing Number
        10,
        // Signal Region
        (-2.208611, 11.807918)
    ).unwrap();
  
    // Deconvoluter with default options
    let deconvoluter: deconvolution::Deconvoluter = Default::default();
  
    // Deconvolute the spectrum
    let deconvoluted_spectrum = deconvoluter.deconvolute_spectrum(&mut spectrum).unwrap();
  
    // WIP for now
}

Here is a simple example of how to use the library in Python:

import matplotlib.pyplot as plt
import metabodecon as md

# Read a spectrum from Bruker TopSpin format
spectrum = md.Spectrum.from_bruker(
    "path/to/spectrum",
    # Experiment Number
    10,
    # Processing Number
    10,
    # Signal Region
    (-2.208611, 11.807918)
)

# Deconvoluter with default options
deconvoluter = md.Deconvoluter()

# Deconvolute the spectrum
deconvolution = deconvoluter.deconvolute_spectrum(spectrum)

# Extract the chemical shifts, intensities, and signal boundaries
x = spectrum.chemical_shifts
y1 = spectrum.intensities_raw
s = spectrum.signal_boundaries

# Compute the superposition of the deconvoluted peaks
y2 = deconvolution.par_superposition_vec(spectrum.chemical_shifts)

# Plot the spectrum with overlayed deconvoluted signal superposition
plt.figure(figsize = (12, 8), dpi = 300)
plt.plot(x, y1, label = "Original Spectrum")
plt.plot(x, y2, label = "Deconvoluted Spectrum", linewidth=0.5)
plt.gca().invert_xaxis()
plt.axvline(x = s[0], color = "black", label = "Signal Boundaries")
plt.axvline(x = s[1], color = "black")
plt.xlabel("Chemical Shifts", fontsize = 16)
plt.ylabel("Intensity", fontsize = 16)
plt.xticks(fontsize = 14)
plt.yticks(fontsize = 14)
plt.legend()
plt.show()

Here is a simple example of how to use the library in R:

# WIP

Developing

Testing

Run tests:

cargo test

Benchmarking

Run benchmarks:

cargo bench

Contributing

Currently not accepting contributions, as this is part of my thesis. However, feel free to use it and open issues for suggestions and bug reports. I will get back to them after my thesis is done.

References

• Hyung-Won Koh et al. “An approach to automated frequency-domain feature extraction in nuclear magnetic resonance spectroscopy”. [DOI] [ScienceDirect]
• Martina Häckl et al. “An R-Package for the Deconvolution and Integration of 1D NMR Data: MetaboDecon1D”. [DOI] [MDPI]

License

This project is licensed under the GPL-3.0 License - see the LICENSE file for details.

Contribution

Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in the work by you, as defined in the GPL-3.0 license, shall be licensed as GPL-3.0, without any additional terms or conditions.