Question: Could tell me how to identify anchor residues in a peptide predicted to bind a particular MHC class I allele?
For HLA class I the main anchors are typically the residues in position 2 and at the C-terminus. There are exceptions (e.g., HLA B*08, and some mouse class I).
In general, though various bioninformatic computational approaches (e.g., neural networks, relative binding matrices, Hidden Markov modeling), class I binding predictions consider the influence of all 20 naturally occurring amino acids on binding capacity. Thus, the final peptide score, usually expressed as a predicted IC50 or percentile score, reflects the peptides entire composition. As such, there is no direct way to look at a prediction and deem specific residues as the anchors.
However, the main anchors may be inferred on the basis of the generality stated above, or by comparing the peptide sequence against allele specific main anchor motifs defined on the basis of pool sequencing and/or peptide elution studies, or as defined empirically by screening peptide libraries etc. The IEDB does calculate summary main anchor motifs for some alleles on the basis of data submitted to the IEDB, and these motifs may be viewed on the IEDB under the Browse tab, and then selecting browse by allele. Other online databases, such as SYFPEITHI, also collect main anchor motifs.