Learning to Classify Organic and Conventional Wheat - A Machine Learning Driven Approach Using the MeltDB 2.0 Metabolomics Analysis Platform

Abstract

We present results of our machine learning approach to the problem of classifying GC-MS data originating from wheat grains of different farming systems. The aim is to investigate the potential of learning algorithms to classify GC-MS data to be either from conventionally grown or from organically grown samples and considering different cultivars. The motivation of our work is rather obvious nowadays: increased demand for organic food in post-industrialized societies and the necessity to prove organic food authenticity. The background of our data set is given by up to 11 wheat cultivars that have been cultivated in both farming systems, organic and conventional, throughout 3 years. More than 300 GC-MS measurements were recorded and subsequently processed and analyzed in the MeltDB 2.0 metabolomics analysis platform, being briefly outlined in this paper. We further describe how unsupervised (t-SNE, PCA) and supervised (SVM) methods can be applied for sample visualization and classification. Our results clearly show that years have most and wheat cultivars have second-most influence on the metabolic composition of a sample. We can also show that for a given year and cultivar, organic and conventional cultivation can be distinguished by machine-learning algorithms.

Keywords: computational metabolomics; food authentication; machine learning; metabolome informatics; metabolomics; organic farming; statistics.

Figure 1

The principal component analysis on the entire dataset of all samples throughout all years, cultivars, and treatments show that the first two components mainly separate samples by the factor year. A separation by the factor farming system is not possible.

Figure 2

A principal component analysis performed on a dataset from 1 year only will mainly cluster samples by their cultivar, regardless of the applied farming system. This PCA is based on samples from the year 2007.

Figure 3

Similar to Figure 1, in the principal component analysis on a dataset of only one cultivar – here “Runal” is shown – the first principal components separate samples by factor year.

Figure 4

Plotting samples from one cultivar (here “Runal”) along the principal components two and four show that a separation by farming system might be possible, even though the main variance is caused by the factor year.

Figure 5

The t-SNE method applied to all samples results in clusters and sub clusters formed according to the factor year and cultivar, respectively.

Figure 6

The same t-SNE result as in Figure 5, but colored by farming system: clusters representing cultivars form subclusters according to the factor farming system.