Sheep milk caseins are promising sources of bioactive peptides with health-enhancing properties. In this study, aS1-, aS2-, ß-, and k-caseins were analyzed using in silico tools assess their physicochemical properties including molecular weight, isoelectric point (pI), instability index, aliphatic index, and GRAVY (Grand Average of Hydropathicity) score as well as their secondary structure composition, post-translational modifications (phosphorylation and glycosylation sites), and conserved sequence motifs. ß-casein showed the highest instability index (100.37) and aliphatic index (89.76), suggesting high flexibility and thermostability. aS2-casein was the most hydrophilic, with the lowest GRAVY score (-1.060) and a balanced charge profile. Secondary structure analysis revealed that ß- and k-caseins are largely disordered, while aS1- and aS2-caseins showed higher alpha-helix content. Simulated gastrointestinal digestion using pepsin, trypsin, and chymotrypsin generated peptide fragments, which were evaluated for bioactivity potential, fragment length, and digestion susceptibility. Antimicrobial activity was predicted using four machine learning classifiers: Random Forest (RF), Support Vector Machine (SVM), Artificial Neural Network (ANN), and Discriminant Analysis (DA). Peptides were further assessed for toxicity, solubility, charge, hydrophobicity, and pI. Machine learning models predicted strong antimicrobial potential for peptides such as PL, AW, PW, and IPIQY. These peptides also exhibited multifunctional activities, including ACE (Angiotensin Converting Enzyme) inhibition and antioxidant efects, and were non-toxic with favorable physicochemical profiles. The findings support the use of sheep milk-derived peptides as natural, multifunctional compounds, contributing to sustainable livestock health management under the One Health framework.