In kimsche/MetaboMate: All you need for NMR-based metabolic profiling with R!

All you need for NMR-based metabolic profiling with R!

---

The R toobox covers all basic processing and analysis steps, including:

• Data Import

  • Fast read-in of 1D NMR spectra in Bruker file format
  • Metadata for each sample
    • Spectrometer-related acquisition parameters (eg. frequency, pulse program, receiver gain)
    • Pre-processing details (eg. time of acquisition, number or data points in time domain, selected AU program)

• Spectral Quality Assessment

  • Detection of baseline distortions
  • Water suppression - quality index
  • Effectiveness shimming / magnetic field homogeneiety using TSP line widths
  • Signal to noise estimation

• Data Pre-Processing

  • Referencing to external standard (TSP) or glucose,
  • Non-linear baseline correction,
  • Easy excision of spectral areas (eg. residual water)

• Sample Normalisation

  • Total area normalisation
  • Probablistic quotient normalisation (PQN)

• Multivariate Analysis and Statistics

  • Principal Comenent Analysis (PCA), including Hotelling's T2 statistic
  • Distance to Model in X space (detection of moderate outliers)
  • Orthogoanal-Patial Least Squares (O-PLS) regression and discriminant analysis
  • Single or multi-column Y for O-PLS

Statistical model interpretation is supported by aestethically pleasing visualisations using ggplot2.

Many of the provided functions can also be applied to mass spectrometry data.

All functions are well-documented, assessible with the help() function in R.

kimsche/MetaboMate documentation built on Aug. 8, 2020, 1:14 a.m.