NCSPHD1238  Preprocessing of quantitative NMR data for Chemometric Analysis

Volume 2014/2015

Quantitative nuclear magnetic resonance (qNMR) is becoming an integral part of many scientific areas, especially within metabolomics, and the handling of the data for the results to be quantitatively reliable is an important issue. This 5-day course has as main objective to introduce tools for pre-processing quantitative NMR data for subsequent multivariate data analysis (chemometrics).

Through lectures and exercises, the course will introduce the participants to important pre-processing methods used within NMR - normalization and alignment (referencing, icoshift) of data - and discuss problems, pitfalls and tricks of the trade in relation to quantitative use of NMR spectroscopy. The course will begin with an introduction to quantitative NMR, including possible artifacts which can occur during acquisition and transformation and possibly spoil quantitative features, and the course will end with the participants doing a principal component analysis (PCA) on their own pre-processed NMR data.
The content of the course is:
Day 1 - Introduction to quantitative NMR, together with Matlab and importing NMR data
Day 2 - Dataset structure, cleaning and Normalization
Day 3 - Alignment 1
Day 4 - Alignment 2 together with basic chemometrics in LatentiX and advanced chemometrics in Matlab
Day 5 - PCA on own NMR dataset and final examination

Learning Outcome


After completing the course the student should be able to:

-How to measure and transform NMR spectra for them to be quantitative
-How to import Bruker NMR spectra into Matlab
-How to install new functions or toolboxes into matlab
-Reflections about raw NMR data (do I need them all?)
-Scripts setup and check (the help function)
-Automatic peak referencing
-Peak alignment: icoshift
-Normalization methods
-Importing processed data into LatentiX and some useful LatentiX tips
-Basic chemometrics in Latentix
-Advanced chemometrics in Matlab: Hints on interval based tools

-Suggest and apply pre-processing tools on quantitative NMR data

-Carry out chemometrics on proper pre-processed quantitative NMR spectra


Handouts and scientific papers provided during the course; scripts and source code provided during the course

Knowledge of chemometrics/exploratory data analysis and basic Matlab beforehand. Basic knowledge of 1H liquid-state NMR spectroscopy.
Lectures and theoretical exercises.
For the project work (doing PCA) the participants should bring their own quantitative NMR data set but if not available a training NMR dataset will be provided.
Type of assessment
Oral examination
In order to pass the course, all participants must perform a PCA on own data and present and discuss the result at an examination on site with the evaluation pass or fail
Marking scale
passed/not passed
Censorship form
No external censorship
  • Category
  • Hours
  • Lectures
  • 20
  • Theory exercises
  • 20
  • Preparation
  • 20
  • Project work
  • 20
  • Total
  • 80