The quantitative binding affinity of 231 9-mer peptides was determined, and used to computationally derive murine TAP specific scoring matrices. The resulting specificity pattern is in good agreement with previous publications that used peptide substitution libraries to directly compare different residues in one position on the affinity or transport rates of murine TAP. In addition, the present study provides the first quantitative specificity data for over seventy residue/position combinations not covered by published substation library data. This complete coverage ensures that no residue/position combination that strongly influences binding has been overlooked, which is necessary to quantitatively predict murine TAP binding specificity for any peptide sequence. We took advantage of being able to compare human and murine TAP specificity matrices, and found that several residues at the Nterminus of peptides that strongly influence binding to human TAP showed little effect on binding to murine TAP. This includes a complete lack of any residue with a strong Fulvestrant positive effect on binding in position 1. In contrast, for peptide C-termini,murine TAP is more specific in its binding preference than human TAP. Taken together, we showed that murine TAP is more skewed than human TAP towards binding peptides based on their C-terminus alone. While not reported as significant in the original publications, examining the figures in references supports such a conclusion. The differences discovered between human and murine TAP binding specificity were shown to correlate with differences in the ability to predict epitope recognition in murine hosts. This demonstrates that our in vitro Gefitinib EGFR/HER2 inhibitor studies correlate with antigen processing events in vivo. It also reinforces that studies of epitope repertoire in mice and human need to take differences between their TAP transporters into account. As TAP is known to transport epitope precursors up to a length of about 16 residues, it is important to characterize its substrate specificity for varying lengths. We were able to successfully predict the affinity of peptides between 8 and 11 residues in length by modeling their binding interaction at the C terminus and the three N terminal residues. In this model of binding, the connecting residues 4 to C-1 of longer peptides are assumed to make only weak interactions with the TAP molecules. This model was previously applied to human TAP, and is shown for the first time to apply to murine TAP as well. The description of murine TAP specificity provides one crucial component towards explaining species specific differences in epitope recognition, which could explain differences in epitope repertoire in humans and HLA transgenic mice frequently used in epitope discovery and vaccine development studies.