Fatty acids-encoded orthogonal partial least squares classifier for authentication of organic beef

Organic beef is a growing market segment due to its well-appreciated health and eco-compatible merits relative to conventional beef produced under mainstream systems. Verifying the authenticity of organic beef is essential to ensure compliance with production regulations and prevent fraudulent mislabeling; however, developing a robust authentication technique is challenging. Here we developed a chemically-informed orthogonal partial least squares classifier, encoding fatty acids (FA) information of intra-muscular fat, with potential for authentication of organic beef. The model successfully achieved a high discriminatory performance between organic and conventional beef (accuracy rates of 99% and 93% for cross and external validation, respectively). Levels of C18:3 n -3, followed by C20:3 n- 3 c 11,14,17, C18:2 n -6, and C22:2 n -6 c 13,16 were the chief discriminant FA that associated with the dietary background of cattle, particularly the amount of forage input. Additionally, levels of C18:2 c 9 t 11, C20:5 n -3 c 5,8,11,14,17, and C22:5 n -3 c 7,10,13,16,19 were significantly higher in organic beef. The approach can contribute to validation of provenance claims associated with organic beef.