As a consequence, various computational tools predicting scores for steps involving the processing an presentation of an MHC binder have been developed, including proteasomal cleavage, TAP transport and MHC affinity. To the author��s knowledge, only two tools exists for predicting the immunogenicty, or the avidity of a potential epitope, namely POPI and POPISK, but are limited to the HLAA2 serotype. Even though these tools are sometimes limited in their accuracy, largely depending on the amount of data they are constructed with, they can provide insights into the mechanism of the emergence of an epitope or the reason for CTL immunity escape. For instance by predicting a lower affinity of a peptide to HLA class I due to the substitution of a favorable residue in an anchor position for a deleterious residue that attenuates or abrogates the ability of a peptide to bind to the particular HLA allotype. In the case of the HLA-B*44 restricted PR 34�C42 epitope EEMNLPGRW, the escape mutation PR D35E is predicted by the MHC binding predictor, NetMHCPan, to have a dramatic negative impact on peptide affinity to HLA�CB*4402, possibly CCG 2046 diminishing the presentation potential of the epitope on HLA�C B*4402. Using statistical methods to infer mutations significantly associated with CTL immunity escape is a challenging task as most public available sequences are rarely annotated with the patient��s HLA genotype. This is especially true of sequences obtained from patients that are not treatment na?��ve. Fortunately, other researchers have found relationships between amino acid substitutions in HIV-1 proteins and HLA allotypes and can potentially be used to provide HLA annotation from HIV-1 protein sequences. Here, it was discovered that there are possible interactions between a common and important protease inhibitor resistance mutation, L90M, and the HLA subtypes B*15, B*48 and potentially A*32. Using the aforementioned data correlating relating amino acid substitutions with HLA subtype, patients were assigned HLA subtypes and the CID 2858522 frequencies of L90M were compared between sequence sets of patients that are HLA B*15/ B*48/A*32 positive and those that are not. It is true that binding affinity does not necessarily constitute an HLA binder to be an epitope, but affinity does improve the stability of the peptide-HLA complex and increases the chance of contact with the appropriate CTL T-Cell Receptor. Inferring a mechanism of escape cannot easily be accomplished computationally. Other researchers have shown in detail general mechanisms of escape that can include diminished interaction of a presented CTL epitope with a complementary cytotoxic T-cell receptor. Another possible method of escape, is the induction of a highly active proteasomal cleavage site within the epitope.