Predicting moisture content during maize nixtamalization using machine learning with NIR spectroscopy

Authors: BURNS, MICHAEL - University Of Minnesota; RENK, JONATHAN - University Of Minnesota; EICKHOLT, DAVID - Pepsico; GILBERT, AMANDA - University Of Minnesota; HATTERY, TRAVIS - Iowa State University; HOLMES, MARK - University Of Minnesota; ANDERSON, NICKOLAS - Pepsico; WATERS, AMANDA - Pepsico; KALAMBUR, SATHYA - Pepsico; Flint-Garcia, Sherry; YANDEAU-NELSON, MARNA - Iowa State University; ANNOR, GEORGE - University Of Minnesota; HIRSCH, CANDICE - University Of Minnesota

Submitted to: Theoretical and Applied Genetics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 7/23/2021
Publication Date: 8/3/2021
Citation: Burns, M.J., Renk, J.S., Eickholt, D.P., Gilbert, A.M., Hattery, T.J., Holmes, M., Anderson, N., Waters, A.J., Kalambur, S., Flint Garcia, S.A., Yandeau-Nelson, M.D., Annor, G.A., Hirsch, C.N. 2021. Predicting moisture content during maize nixtamalization using machine learning with NIR spectroscopy. Theoretical and Applied Genetics. 134:3743–3757. https://doi.org/10.1007/s00122-021-03926-8.
DOI: https://doi.org/10.1007/s00122-021-03926-8

Interpretive Summary: Many food products made from corn, including tortillas and corn chips, rely on a cooking process called nixtamalization. During nixtamalization, the grain is cooked in an alkaline solution that removes the indigestible outer layer of the kernel, softens the grain to facilitate grinding to make masa dough, contributes to the texture of the final food product, and makes vitamins and other nutrients more available during digestion. One of the key variables that determines the outcome of nixtamalization is moisture content of the grain during and after cooking. However, measuring moisture content during nixtamalization is time consuming and low throughput, which is not amenable to the large number of samples required in a corn breeding program focused on improving food-grade corn. Therefore, a method to rapidly predict moisture content during nixtamalization is necessary to make improvements in food-grade corn germplasm that is sourced for masa-based products. In this study, we collected near-infrared (NIR) spectroscopy data for a large number of corn samples adapted to the US Corn Belt and quantified moisture content during nixtamalization on the same grain samples. Using these data, we developed statistical models that can accurately predict moisture content from the NIR spectra without the need to conduct nixtamalization. We also conducted a statistical analysis to find genes involved in moisture content during nixtamalization, and determined that this trait is controlled by many genes. Our results provide a high-throughput method to evaluate moisture content during nixtamalization that is feasible at the scale of a breeding program, and provides important information about the factors contributing to variation for this trait to breeders. Manufacturers of masa-based products may also be interested in incorporating this system into their pipelines to predictively alter nixtamalization conditions.

Technical Abstract: Lack of high throughput, high-quality phenotyping systems for determining moisture content during the maize nixtamalization cooking process has led to difficulty in breeding for this trait. This study provides a high throughput, quantitative measure of kernel moisture content during nixtamalization based on NIR scanning of uncooked maize kernels. Machine learning was utilized to develop models based on the combination of NIR spectra and moisture content determined from a scaled-down benchtop cook method. A linear support vector machine (SVM) model with a Spearman’s rank correlation coefficient of 0.852 between wet laboratory and predicted values was developed from 100 diverse temperate genotypes grown in replicate across two environments. This model was applied to NIR spectra data from 501 diverse temperate genotypes grown in replicate in five environments. Analysis of variance revealed environment explained the highest percent of the variation (51.5%), followed by genotype (15.6%) and genotype-by-environment interaction (11.2%). A genome-wide association study identified 26 significant loci across five environments that explained between 5.04% and 16.01% (average = 10.41%). However, genome-wide markers explained 10.54% to 45.99% (average = 31.68%) of the variation, indicating the genetic architecture of this trait is likely complex and controlled by many loci of small effect. This study provides a high-throughput method to evaluate moisture content during nixtamalization that is feasible at the scale of a breeding program and provides important information about the factors contributing to variation of this trait for breeders and food companies to make future strategies to improve this important processing trait.