A series of experiments was conducted to eliminate activity of SBA. Deglycosylation decreased activity of SBA by 21%, but not as much as denaturation (23-53%). Single enzymes did not hydrolyze native SBA, but hydrolyzed heat- or organic solutes-denatured SBA. Even after hydrolysis, activity of SBA still was not fully eliminated (44-62% residual). A combination of multiple enzymes with thermolysin fully deactivated heat- or organic solutes-treated SBA. Tea polyphenols, metal ions, chelating agents were also tested and had no significant effect on deactivating SBA. N-acetyl-galactosamine (GalNAc)-agrose beads specifically removed SBA from a protein mixture, but not fully and activity of SBA was not eliminated. Pepsin and pancreatin hydrolysis fully deactivated native SBA. This treatment uses minimum heating, as a result, it was chosen to treat soy white flake (SWF) for feeding trial to evaluate feeding quality. The activity of beta-conglycinin was not reduced by single enzyme hydrolysis, but was fully eliminated by a combination of multiple enzyme treatments. Activity of glycinin was not fully reduced by either single or a combination of multiple enzyme hydrolysis. Similarly, the activities of SWF and soy protein isolate (SPI) were not fully reduced by single or combination of multiple enzyme hydrolysis. Although pepsin and pancreatin treatment did not fully deactivate beta-conglycinin, glycinin, SWF and SPI, this method was used to generate feed material for in vivo evaluation of nutritional quality due to the ability to fully deactivate SBA. The in vitro study showed that the SBA in SWF was deactivated by this treatment, but did not improve chick growth performance compared to the raw SWF, however, the chicks did not show any pancreas enlargement or intestine weight increase compared to the raw soy feeding or a commercial diet, indicating the deactivation of SBA in the material. After enzyme hydrolysis, SBA, beta-conglycinin and glycinin still had hemagglutination activity, this result caused us to hypothesize that certain peptides in these proteins had hemagglutination activity. In order to identify the active peptides, the three proteins were subjected to trypsin hydrolysis, and GalNAc-agrose beads were used to isolate the active peptides. MALDI-TOF (Matrix Assisted Laser Desorption/Ionization Time-of-Flight) was used to identify the masses of the peptides, and the masses were compared to the peptide profiles given by theoretical cleavage of the proteins, so that the peptides could be identified. Two peptides from SBA, 24 peptides from beta-conglycinin and 16 peptides from glycinin were identified from the active peptide extracts. In addition, 2, 3 and 3 peptides from SBA, beta-conglycinin and glycinin respectively were synthesized and their activities were assessed using hemagglutination assay. These peptides have hemagglutination activity whereas a synthesized control peptide from SBA did not show any activity. This result confirmed our hypothesis that there are active peptides in soy protein that have hemagglutination activity.