Limits...
MultiRTA: a simple yet reliable method for predicting peptide binding affinities for multiple class II MHC allotypes.

Bordner AJ, Mittelmann HD - BMC Bioinformatics (2010)

Bottom Line: Comparison of the HLA-DR results with those of two other prediction methods applied to the same data sets showed that MultiRTA achieved performance comparable to NetMHCIIpan and better than the earlier TEPITOPE method.An examination of predictions for the promiscuous binding CLIP peptide revealed variations in binding affinity among alleles as well as potentially distinct binding registers for HLA-DR and HLA-DP.The MultiRTA method yields competitive performance but with a significantly simpler and physically interpretable model compared with previous prediction methods.

View Article: PubMed Central - HTML - PubMed

Affiliation: Mayo Clinic, 13400 East Shea Boulevard, Scottsdale, AZ 85259, USA. bordner.andrew@mayo.edu

ABSTRACT

Background: The binding of peptide fragments of antigens to class II MHC is a crucial step in initiating a helper T cell immune response. The identification of such peptide epitopes has potential applications in vaccine design and in better understanding autoimmune diseases and allergies. However, comprehensive experimental determination of peptide-MHC binding affinities is infeasible due to MHC diversity and the large number of possible peptide sequences. Computational methods trained on the limited experimental binding data can address this challenge. We present the MultiRTA method, an extension of our previous single-type RTA prediction method, which allows the prediction of peptide binding affinities for multiple MHC allotypes not used to train the model. Thus predictions can be made for many MHC allotypes for which experimental binding data is unavailable.

Results: We fit MultiRTA models for both HLA-DR and HLA-DP using large experimental binding data sets. The performance in predicting binding affinities for novel MHC allotypes, not in the training set, was tested in two different ways. First, we performed leave-one-allele-out cross-validation, in which predictions are made for one allotype using a model fit to binding data for the remaining MHC allotypes. Comparison of the HLA-DR results with those of two other prediction methods applied to the same data sets showed that MultiRTA achieved performance comparable to NetMHCIIpan and better than the earlier TEPITOPE method. We also directly tested model transferability by making leave-one-allele-out predictions for additional experimentally characterized sets of overlapping peptide epitopes binding to multiple MHC allotypes. In addition, we determined the applicability of prediction methods like MultiRTA to other MHC allotypes by examining the degree of MHC variation accounted for in the training set. An examination of predictions for the promiscuous binding CLIP peptide revealed variations in binding affinity among alleles as well as potentially distinct binding registers for HLA-DR and HLA-DP. Finally, we analyzed the optimal MultiRTA parameters to discover the most important peptide residues for promiscuous and allele-specific binding to HLA-DR and HLA-DP allotypes.

Conclusions: The MultiRTA method yields competitive performance but with a significantly simpler and physically interpretable model compared with previous prediction methods. A MultiRTA prediction webserver is available at http://bordnerlab.org/MultiRTA.

Show MeSH

Related in: MedlinePlus

Variation in the peptide core residue binding specificity as assessed by the standard deviation in MultiRTA predicted binding affinity over all 20 residue types. Each bar represents the variation for one MHC allotype in the training data set.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2957400&req=5

Figure 1: Variation in the peptide core residue binding specificity as assessed by the standard deviation in MultiRTA predicted binding affinity over all 20 residue types. Each bar represents the variation for one MHC allotype in the training data set.

Mentions: We also examined the relative importance of each of the nine peptide core residues in binding specificity. This was done by calculating the standard deviation over all 20 residue types at core residue Pi for each residue type T in the training sets, σT, i defined in Eq. 11. These values are plotted for HLA-DR and HLA-DP in Figures 1(a) and 1(b), respectively. We also calculated the difference between the maximum and minimum binding affinity values and obtained qualitatively similar results (data not shown). For HLA-DR, it may be seen that P1 has the largest variation and so makes the largest contribution to binding specificity. This is consistent with the previous identification of P1 as a primary anchor residue for HLA-DR and with the discussion above. Peptide residues P4, P6, P7, and P9 also make large contributions to specificity relative to the remaining residues. Residue P4, P6, and P9 have been previously classified as secondary anchors on the basis of their contribution to binding specificity and the fact that they, along with P1 bind into four pockets in the MHC binding cleft. It is interesting that other peptide core residues, especially P7, also make significant contributions to peptide binding specificity. All nine peptide core side chains contact the MHC in some X-ray structure and so can potentially form energetically favorable interactions that stabilize peptide binding. Compared with peptide-class I MHC complexes, peptides bound to class II MHC assume an extended conformation low in the binding cleft rather than bulging outside of the binding cleft and so can form more extensive contacts with the MHC protein [46]. This also explains the general lack of well-defined anchor residues for class II MHC allotypes as compared with class I MHC. Thus, peptide core residues outside of the traditional anchor residues also make significant contributions to peptide binding to HLA-DR.


MultiRTA: a simple yet reliable method for predicting peptide binding affinities for multiple class II MHC allotypes.

Bordner AJ, Mittelmann HD - BMC Bioinformatics (2010)

Variation in the peptide core residue binding specificity as assessed by the standard deviation in MultiRTA predicted binding affinity over all 20 residue types. Each bar represents the variation for one MHC allotype in the training data set.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2957400&req=5

Figure 1: Variation in the peptide core residue binding specificity as assessed by the standard deviation in MultiRTA predicted binding affinity over all 20 residue types. Each bar represents the variation for one MHC allotype in the training data set.
Mentions: We also examined the relative importance of each of the nine peptide core residues in binding specificity. This was done by calculating the standard deviation over all 20 residue types at core residue Pi for each residue type T in the training sets, σT, i defined in Eq. 11. These values are plotted for HLA-DR and HLA-DP in Figures 1(a) and 1(b), respectively. We also calculated the difference between the maximum and minimum binding affinity values and obtained qualitatively similar results (data not shown). For HLA-DR, it may be seen that P1 has the largest variation and so makes the largest contribution to binding specificity. This is consistent with the previous identification of P1 as a primary anchor residue for HLA-DR and with the discussion above. Peptide residues P4, P6, P7, and P9 also make large contributions to specificity relative to the remaining residues. Residue P4, P6, and P9 have been previously classified as secondary anchors on the basis of their contribution to binding specificity and the fact that they, along with P1 bind into four pockets in the MHC binding cleft. It is interesting that other peptide core residues, especially P7, also make significant contributions to peptide binding specificity. All nine peptide core side chains contact the MHC in some X-ray structure and so can potentially form energetically favorable interactions that stabilize peptide binding. Compared with peptide-class I MHC complexes, peptides bound to class II MHC assume an extended conformation low in the binding cleft rather than bulging outside of the binding cleft and so can form more extensive contacts with the MHC protein [46]. This also explains the general lack of well-defined anchor residues for class II MHC allotypes as compared with class I MHC. Thus, peptide core residues outside of the traditional anchor residues also make significant contributions to peptide binding to HLA-DR.

Bottom Line: Comparison of the HLA-DR results with those of two other prediction methods applied to the same data sets showed that MultiRTA achieved performance comparable to NetMHCIIpan and better than the earlier TEPITOPE method.An examination of predictions for the promiscuous binding CLIP peptide revealed variations in binding affinity among alleles as well as potentially distinct binding registers for HLA-DR and HLA-DP.The MultiRTA method yields competitive performance but with a significantly simpler and physically interpretable model compared with previous prediction methods.

View Article: PubMed Central - HTML - PubMed

Affiliation: Mayo Clinic, 13400 East Shea Boulevard, Scottsdale, AZ 85259, USA. bordner.andrew@mayo.edu

ABSTRACT

Background: The binding of peptide fragments of antigens to class II MHC is a crucial step in initiating a helper T cell immune response. The identification of such peptide epitopes has potential applications in vaccine design and in better understanding autoimmune diseases and allergies. However, comprehensive experimental determination of peptide-MHC binding affinities is infeasible due to MHC diversity and the large number of possible peptide sequences. Computational methods trained on the limited experimental binding data can address this challenge. We present the MultiRTA method, an extension of our previous single-type RTA prediction method, which allows the prediction of peptide binding affinities for multiple MHC allotypes not used to train the model. Thus predictions can be made for many MHC allotypes for which experimental binding data is unavailable.

Results: We fit MultiRTA models for both HLA-DR and HLA-DP using large experimental binding data sets. The performance in predicting binding affinities for novel MHC allotypes, not in the training set, was tested in two different ways. First, we performed leave-one-allele-out cross-validation, in which predictions are made for one allotype using a model fit to binding data for the remaining MHC allotypes. Comparison of the HLA-DR results with those of two other prediction methods applied to the same data sets showed that MultiRTA achieved performance comparable to NetMHCIIpan and better than the earlier TEPITOPE method. We also directly tested model transferability by making leave-one-allele-out predictions for additional experimentally characterized sets of overlapping peptide epitopes binding to multiple MHC allotypes. In addition, we determined the applicability of prediction methods like MultiRTA to other MHC allotypes by examining the degree of MHC variation accounted for in the training set. An examination of predictions for the promiscuous binding CLIP peptide revealed variations in binding affinity among alleles as well as potentially distinct binding registers for HLA-DR and HLA-DP. Finally, we analyzed the optimal MultiRTA parameters to discover the most important peptide residues for promiscuous and allele-specific binding to HLA-DR and HLA-DP allotypes.

Conclusions: The MultiRTA method yields competitive performance but with a significantly simpler and physically interpretable model compared with previous prediction methods. A MultiRTA prediction webserver is available at http://bordnerlab.org/MultiRTA.

Show MeSH
Related in: MedlinePlus