An integrated in silico, in vitro, and machine learning pipeline for rapid discovery of antioxidant peptides and co-product lipids from goat liver

Abstract

Goat liver abundant in high-quality protein is a by-product during meat processing, which is currently underutilized. Traditional enzymatic hydrolysis and screening methods suffer from low extraction efficiency and unstable yields. This study aims to develop an integrated pipeline for rapid discovery of antioxidant peptides co-product lipids from goat liver. Papain was selected as the optimal protease, with hydrolysis time of 5 h. Peptidomics and machine learning enabled identification and activity-based prediction of antioxidant peptides, leading to discovery of two novel antioxidant peptides (WGF and GPLF) with IC₅₀ values for DPPH radical scavenging activities of 1244 μM and 2534 μM, respectively. Molecular docking and dynamics simulations indicated the stable binding of both peptides to Keap1 via hydrogen bond, van der Waals force, and water-mediated interactions, suggesting potential antioxidant activity through Keap1-Nrf2 pathway. GC–MS analysis revealed that saturated and unsaturated fatty acids accounted for 25.33 ± 0.02% and 74.67 ± 0.04%, respectively. This study develops an integrated pipeline for extraction of antioxidant peptides and co-product lipids from goat liver, and contributes to the future development of bioactive peptide-based packaging systems.