Integrated modeling of the major events in the MHC class I antigen processing pathway

Rational design of epitope-driven vaccines is a key goal of immunoinformatics. Typically, candidate selection relies on the prediction of MHC-peptide binding only, as this is known to be the most selective step in the MHC class I antigen processing pathway. However, proteasomal cleavage and transport by the transporter associated with antigen processing (TAP) are essential steps in antigen processing as well. While prediction methods exist for the individual steps, no method has yet offered an integrated prediction of all three major processing events. Here we present WAPP, a method combining prediction of proteasomal cleavage, TAP transport, and MHC binding into a single prediction system. The proteasomal cleavage site prediction employs a new matrix-based method that is based on experimentally verified proteasomal cleavage sites. Support vector regression is used for predicting peptides transported by TAP. MHC binding is the last step in the antigen processing pathway and was predicted using a support vector machine method, SVMHC. The individual methods are combined in a filtering approach mimicking the natural processing pathway. WAPP thus predicts peptides that are cleaved by the proteasome at the C terminus, transported by TAP, and show significant affinity to MHC class I molecules. This results in a decrease in false positive rates compared to MHC binding prediction alone. Compared to prediction of MHC binding only, we report an increased overall accuracy and a lower rate of false positive predictions for the HLA-A*0201, HLA-B*2705, HLA-A*01, and HLA-A*03 alleles using WAPP. The method is available online through our prediction server at http://www-bs.informatik.uni-tuebingen.de/WAPP     

Towards the in silico identification of class II restricted T-cell epitopes: a partial least squares iterative self-consistent algorithm for affinity prediction

MOTIVATION: The immunogenicity of peptides depends on their ability to bind to MHC molecules. MHC binding affinity prediction methods can save significant amounts of experimental work. The class II MHC binding site is open at both ends, making epitope prediction difficult because of the multiple binding ability of long peptides. RESULTS: An iterative self-consistent partial least squares (PLS)-based additive method was applied to a set of 66 peptides no longer than 16 amino acids, binding to DRB1*0401. A regression equation containing the quantitative contributions of the amino acids at each of the nine positions was generated. Its predictability was tested using two external test sets which gave r(pred) = 0.593 and r(pred) = 0.655, respectively. Furthermore, it was benchmarked using 25 known T-cell epitopes restricted by DRB1*0401 and we compared our results with four other online predictive methods. The additive method showed the best result finding 24 of the 25 T-cell epitopes. AVAILABILITY: Peptides used in the study are available from http://www.jenner.ac.uk/JenPep. The PLS method is available commercially in the SYBYL molecular modelling software package. The final model for affinity prediction of peptides binding to DRB1*0401 molecule is available at http://www.jenner.ac.uk/MHCPred. Models developed for DRB1*0101 and DRB1*0701 also are available in MHCPred.     

Exploring the minimal functional unit of the transporter associated with antigen processing

TAP, an ABC transporter in the ER membrane, provides antigenic peptides derived from proteasomal degradation to MHC class I molecules for inspection by cytotoxic T lymphocytes at the cell surface so as to trace malignant or infected cells. To investigate the minimal number of transmembrane segments (TMs) required for assembly of the TAP complex based on hydrophobicity algorithms and alignments with other ABC transporters we generated N-terminal truncation variants of human TAP1 and TAP2. As a result, a 6+6 TM core-TAP complex represents the minimal functional unit of the transporter, which is essential and sufficient for heterodimer assembly, peptide binding, and peptide translocation into the ER. The TM1 of both, core-TAP1 and core-TAP2 are critical for heterodimerization of the complex.     

Peptides induce ATP hydrolysis at both subunits of the transporter associated with antigen processing

The transporter associated with antigen processing (TAP) plays a key role in the adaptive immune response by pumping antigenic peptides into the endoplasmic reticulum for subsequent loading of major histocompatibility complex class I molecules. TAP is a heterodimer consisting of TAP1 and TAP2. Each subunit is composed of a transmembrane domain and a nucleotide-binding domain, which energizes the peptide transport. To analyze ATP hydrolysis of each subunit we developed a method of trapping 8-azido-nucleotides to TAP in the presence of phosphate transition state analogs followed by photocross-linking, immunoprecipitation, and high resolution SDS-PAGE. Strikingly, trapping of both TAP subunits by beryllium fluoride is peptide-specific. The peptide concentration required for half-maximal trapping is identical for TAP1 and TAP2 and directly correlates with the peptide binding affinity. Only a background level of trapping was observed for low affinity peptides or in the presence of the herpes simplex viral protein ICP47, which specifically blocks peptide binding to TAP. Importantly, the peptide-induced trapped state is reached after ATP hydrolysis and not in a backward reaction of ADP binding and trapping. In the trapped state, TAP can neither bind nor exchange nucleotides, whereas peptide binding is not affected. In summary, these data support the model that peptide binding induces a conformation that triggers ATP hydrolysis in both subunits of the TAP complex within the catalytic cycle.     

Peptide-bound major histocompatibility complex class I molecules associate with tapasin before dissociation from transporter associated with antigen processing

Major histocompatibility complex (MHC) class I molecules present antigenic peptides to CD8 T cells. The peptides are generated in the cytosol, then translocated across the membrane of the endoplasmic reticulum by the transporter associated with antigen processing (TAP). TAP is a trimeric complex consisting of TAP1, TAP2, and tapasin (TAP-A) as indicated for human cells by reciprocal coprecipitation with anti-TAP1/2 and anti-tapasin antibodies, respectively. TAP1 and TAP2 are required for the peptide transport. Tapasin is involved in the association of class I with TAP and in the assembly of class I with peptide. The mechanisms of tapasin function are still unknown. Moreover, there has been no evidence for a murine tapasin analogue, which has led to the suggestion that murine MHC class I binds directly to TAP1/2. In this study, we have cloned the mouse analogue of tapasin. The predicted amino acid sequence showed 78% identity to human tapasin with identical consensus sequences of signal peptide, N-linked glycosylation site, transmembrane domain and double lysine motif. However, there was less homology (47%) found at the predicted cytosolic domain, and in addition, mouse tapasin is 14 amino acids longer than the human analogue at the C terminus. This part of the molecule may determine the species specificity for interaction with MHC class I or TAP1/2. Like human tapasin, mouse tapasin binds both to TAP1/2 and MHC class I. In TAP2-mutated RMA-S cells, both TAP1 and MHC class I were coprecipitated by anti-tapasin antiserum indicative of association of tapasin with TAP1 but not TAP2. With crosslinker-modified peptides and purified microsomes, anti-tapasin coprecipitated both peptide-bound MHC class I and TAP1/2. In contrast, anti-calreticulin only coprecipitated peptide-free MHC class I molecules. This difference in association with peptide-loaded class I suggests that tapasin functions later than calreticulin during MHC class I assembly, and controls peptide loading onto MHC class I molecules in the endoplasmic reticulum.     

Better prediction of the location of alpha-turns in proteins with support vector machine

We have developed a novel method named AlphaTurn to predict alpha-turns in proteins based on the support vector machine (SVM). The prediction was done on a data set of 469 nonhomologous proteins containing 967 alpha-turns. A great improvement in prediction performance was achieved by using multiple sequence alignment generated by PSI-BLAST as input instead of the single amino acid sequence. The introduction of secondary structure information predicted by PSIPRED also improved the prediction performance. Moreover, we handled the very uneven data set by combining the cost factor j with the "state-shifting" rule. This further promoted the prediction quality of our method. The final SVM model yielded a Matthews correlation coefficient (MCC) of 0.25 by a 10-fold cross-validation. To our knowledge, this MCC value is the highest obtained so far for predicting alpha-turns. An online Web server based on this method has been developed and can be freely accessed at http://bmc.hust.edu.cn/bioinformatics/ or http://210.42.106.80/.     

Localization, expression change in PRRSV infection and association analysis of the porcine TAP1 gene

The transporter associated with antigen processing (TAP) translocates antigenic peptides from the cytosol into the lumen of the endoplasmic reticular and plays a critical role in the major histocompatibility complex (MHC) class I molecule-mediated antigenic presentation pathway. In this study, the porcine TAP1 gene was mapped to the pig chromosome 7 (SSC7) and was closely linked to the marker SSC2B02 (retention fraction=43%, LOD=15.18). Subcellular localization of TAP1 by transient transfection of PK15 cells indicated that the TAP1 protein might be located in the endoplasmic reticulum (ER) in pig kidney epithelial cells (PK-15). Gene expression analysis by semi-quantitative RT-PCR revealed that TAP1 was selectively expressed in some immune and immune-related tissues. Quantitative real-time PCR (qRT-PCR) analysis revealed that this gene was up-regulated after treatments that mimic viral and bacterial infection (polyriboinosinic-polyribocytidylic acid (poly(I:C)) and lipopolysaccharide (LPS), respectively). In addition, elevated TAP1 expression was detected after porcine reproductive and respiratory syndrome virus (PRRSV) infection in porcine white blood cells (WBCs). One single nucleotide polymorphism (SNP) in exon 3 of TAP1 was detected in a Landrace pig population by Bsp143I restriction enzyme digestion. Different genotypes of this SNP had significant associations (P<0.05) with the red blood cell distribution width (RDW) of 1-day-old (1 d) pigs (P=0.0168), the PRRSV antibody level (PRRSV Ab) (P=0.0445) and the absolute lymphocyte count (LYM#) (P=0.024) of 17 d pigs. Our results showed that the TAP1 gene might have important roles in swine immune responses, and these results provide useful information for further functional studies.     

A neural network model approach to the study of human TAP transporter

We used an artificial neural network (ANN) computer model to study peptide binding to the human transporter associated with antigen processing (TAP). After validation, an ANN model of TAP-peptide binding was used to mine a database of HLA-binding peptides to elucidate patterns of TAP binding. The affinity of HLA-binding peptides for TAP was found to differ according to the HLA supertype concerned: HLA-B27, -A3 or -A24 binding peptides had high, whereas HLA-A2, -B7 or -B8 binding peptides had low affinity for TAP. These results support the idea that TAP and particular HLA molecules may have co-evolved for efficient peptide processing and presentation. The strong similarity between the sets of peptides bound by TAP or HLA-B27 suggests functional co-evolution whereas the lack of a relationship between the sets of peptides bound by TAP or HLA-A2 is against these particular molecules having co-evolved. In support of these conclusions, the affinities of HLA-A2 and HLA-B7 binding peptides for TAP show similar distributions to that of randomly generated peptides. On the basis of these results we propose that HLA alleles constitute two separate classes: those that are TAP-efficient for peptide loading (HLA-B27, -A3 and -A24) and those that are TAP-inefficient (HLA-A2, -B7 and -B8). Computer modelling can be used to complement laboratory experiments and thereby speed up knowledge discovery in biology. In particular, we provide evidence that large-scale experiments can be avoided by combining initial experimental data with limited laboratory experiments sufficient to develop and validate appropriate computer models. These models can then be used to perform large-scale simulated experiments the results of which can then be validated by further small-scale laboratory experiments.     

A Viral,Transporter Associated with Antigen Processing (TAP)-independent,High Affinity Ligand with Alternative Interactions Endogenously Presented by the Nonclassical Human Leukocyte Antigen E Class I Molecule

The transporter associated with antigen processing (TAP) enables the flow of viral peptides generated in the cytosol by the proteasome and other proteases to the endoplasmic reticulum, where they complex with nascent human leukocyte antigen (HLA) class I. Later, these peptide-HLA class I complexes can be recognized by CD8⁺ lymphocytes. Cancerous cells and infected cells in which TAP is blocked, as well as individuals with unusable TAP complexes, are able to present peptides on HLA class I by generating them through TAP-independent processing pathways. Here, we identify a physiologically processed HLA-E ligand derived from the D8L protein in TAP-deficient vaccinia virus-infected cells. This natural high affinity HLA-E class I ligand uses alternative interactions to the anchor motifs previously described to be presented on nonclassical HLA class I molecules. This octameric peptide was also presented on HLA-Cw1 with similar binding affinity on both classical and nonclassical class I molecules. In addition, this viral peptide inhibits HLA-E-mediated cytolysis by natural killer cells. Comparison between the amino acid sequences of the presenting HLA-E and HLA-Cw1 alleles revealed a shared structural motif in both HLA class molecules, which could be related to their observed similar cross-reactivity affinities. This motif consists of several residues located on the floor of the peptide-binding site. These data expand the role of HLA-E as an antigen-presenting molecule.     

Disruption of the association of 73 kDa heat shock cognate protein with transporters associated with antigen processing (TAP) decreases TAP-dependent translocation of antigenic peptides into the endoplasmic reticulum

Major histocompatibility complex class I-bound antigenic peptides generated in the cytosol are translocated into the ER by TAP. In the present study, the physical association of HSC73 with TAP in human lymphoblastoid T1 cells was demonstrated. The dissociation was induced in the presence of 10 mM ATP, indicating that the ADP-binding form of HSC73 might be associated with TAP. We found that HSC73-binding immunosuppressant, MeDSG disrupted the HSC73-TAP association, whereas it did not affect the binding of HSC73 to a substrate protein. MHC class I expression on the cell surface was also downregulated. Then, the effect of MeDSG on the TAP-mediated ER translocation was examined using two homologous model peptides, NGT-Bw4 and NGT-Bw6, which had distinct binding affinity to HSC73. Although high-affinity peptide NGT-Bw4 was translocated by TAP, low-affinity peptide NGT-Bw6 was not. The TAP-dependent translocation of NGT-Bw4 was abolished in the presence of MeDSG. Decreased presentation on the cell surface was shown for the human leukocyte antigen (HLA)-A31-restricted natural antigenic peptide F4.2, which had high affinity to HSC73, in the presence of MeDSG. It was indicated that disruption of the HSC73-TAP association resulted in inhibition of TAP-dependent translocation of HSC73-bound peptides. Our findings highlighted an important role of HSC73 for feeding antigenic peptides to TAP, and suggested a possibility that a synthetic polyamine might inhibit the function of HSC73, thereby suppressing MHC class I-restricted presentation of HSC73-bound antigenic peptides.     

Molecular mechanism and species specificity of TAP inhibition by herpes simplex virus ICP47.

The immediate early protein ICP47 of herpes simplex virus (HSV) inhibits the transporter for antigen processing (TAP)-mediated translocation of antigen-derived peptides across the endoplasmic reticulum (ER) membrane. This interference prevents assembly of peptides with class I MHC molecules in the ER and ultimately recognition of HSV-infected cells by cytotoxic T-lymphocytes, potentially leading to immune evasion of the virus. Here, we demonstrate that recombinant, purified ICP47 containing a hexahistidine tag inhibits peptide import into microsomes of insect cells expressing human TAP, whereas inhibition of peptide transport by murine TAP was much less effective. This finding indicates an intrinsic species-specificity of ICP47 and suggests that no additional proteins interacting specifically with either ICP47 or TAP are required for inhibition of peptide transport. Since neither purified nor induced ICP47 inhibited photocrosslinking of 8-azido-ATP to TAP1 and TAP2 it seems that ICP47 does not prevent ATP from binding to TAP. By contrast, peptide binding was completely blocked by ICP47 as shown both by photoaffinity crosslinking of peptides to TAP and peptide binding to microsomes from TAP-transfected insect cells. Competition experiments indicated that ICP47 binds to human TAP with a higher affinity (50 nM) than peptides whereas the affinity to murine TAP was 100-fold lower. Our data suggest that ICP47 prevents peptides from being translocated by blocking their binding to the substrate-binding site of TAP.     

Analyses of Conformational States of the Transporter Associated with Antigen Processing (TAP) Protein in a Native Cellular Membrane Environment

The transporter associated with antigen processing (TAP) plays a critical role in the MHC class I antigen presentation pathway. TAP translocates cellular peptides across the endoplasmic reticulum membrane in an ATP hydrolysis-dependent manner. We used FRET spectroscopy in permeabilized cells to delineate different conformational states of TAP in a native subcellular membrane environment. For these studies, we tagged the TAP1 and TAP2 subunits with enhanced cyan fluorescent protein and enhanced yellow fluorescent protein, respectively, C-terminally to their nucleotide binding domains (NBDs), and measured FRET efficiencies under different conditions. Our data indicate that both ATP and ADP enhance the FRET efficiencies but that neither induces a maximally closed NBD conformation. Additionally, peptide binding induces a large and significant increase in NBD proximity with a concentration dependence that is reflective of individual peptide affinities for TAP, revealing the underlying mechanism of peptide-stimulated ATPase activity of TAP. Maximal NBD closure is induced by the combination of peptide and non-hydrolysable ATP analogs. Thus, TAP1-TAP2 NBD dimers are not fully stabilized by nucleotides alone, and substrate binding plays a key role in inducing the transition state conformations of the NBD. Taken together, these findings show that at least three steps are involved in the transport of peptides across the endoplasmic reticulum membrane for antigen presentation, corresponding to three dynamically and structurally distinct conformational states of TAP. Our studies elucidate structural changes in the TAP NBD in response to nucleotides and substrate, providing new insights into the mechanism of ATP-binding cassette transporter function.     

POPI: predicting immunogenicity of MHC class I binding peptides by mining informative physicochemical properties

MOTIVATION: Both modeling of antigen-processing pathway including major histocompatibility complex (MHC) binding and immunogenicity prediction of those MHC-binding peptides are essential to develop a computer-aided system of peptide-based vaccine design that is one goal of immunoinformatics. Numerous studies have dealt with modeling the immunogenic pathway but not the intractable problem of immunogenicity prediction due to complex effects of many intrinsic and extrinsic factors. Moderate affinity of the MHC-peptide complex is essential to induce immune responses, but the relationship between the affinity and peptide immunogenicity is too weak to use for predicting immunogenicity. This study focuses on mining informative physicochemical properties from known experimental immunogenicity data to understand immune responses and predict immunogenicity of MHC-binding peptides accurately. RESULTS: This study proposes a computational method to mine a feature set of informative physicochemical properties from MHC class I binding peptides to design a support vector machine (SVM) based system (named POPI) for the prediction of peptide immunogenicity. High performance of POPI arises mainly from an inheritable bi-objective genetic algorithm, which aims to automatically determine the best number m out of 531 physicochemical properties, identify these m properties and tune SVM parameters simultaneously. The dataset consisting of 428 human MHC class I binding peptides belonging to four classes of immunogenicity was established from MHCPEP, a database of MHC-binding peptides (Brusic et al., 1998). POPI, utilizing the m = 23 selected properties, performs well with the accuracy of 64.72% using leave-one-out cross-validation, compared with two sequence alignment-based prediction methods ALIGN (54.91%) and PSI-BLAST (53.23%). POPI is the first computational system for prediction of peptide immunogenicity based on physicochemical properties. AVAILABILITY: A web server for prediction of peptide immunogenicity (POPI) and the used dataset of MHC class I binding peptides (PEPMHCI) are available at http://iclab.life.nctu.edu.tw/POPI     

Structural and Functional Dissection of the Human Cytomegalovirus Immune Evasion Protein US6

The human cytomegalovirus (HCMV) protein US6 inhibits the transporter associated with antigen processing (TAP). Since TAP transports antigenic peptides into the endoplasmic reticulum for binding to major histocompatibility class I molecules, inhibition of the transporter by HCMV US6 impairs the presentation of viral antigens to cytotoxic T lymphocytes. HCMV US6 inhibits ATP binding by TAP, hence depriving TAP of the energy source it requires for peptide translocation, yet the molecular basis for the interaction between US6 and TAP is poorly understood. In this study we demonstrate that residues 89 to 108 of the HCMV US6 luminal domain are required for TAP inhibition, whereas sequences that flank this region stabilize the binding of the viral protein to TAP. In parallel, we demonstrate that chimpanzee cytomegalovirus (CCMV) US6 binds, but does not inhibit, human TAP. The sequence of CCMV US6 differs from that of HCMV US6 in the region corresponding to residues 89 to 108 of the HCMV protein. The substitution of this region of CCMV US6 with the corresponding residues from HCMV US6 generates a chimeric protein that inhibits human TAP and provides further evidence for the pivotal role of residues 89 to 108 of HCMV US6 in the inhibition of TAP. On the basis of these observations, we propose that there is a hierarchy of interactions between HCMV US6 and TAP, in which residues 89 to 108 of HCMV US6 interact with and inhibit TAP, whereas other parts of the viral protein also bind to TAP and stabilize this inhibitory interaction.     

Modulation of the antigenic peptide transporter TAP by recombinant antibodies binding to the last five residues of TAP1

The transporter associated with antigen processing (TAP) plays a pivotal role in the major histocompatibility complex (MHC) class I mediated immune response against infected or malignantly transformed cells. It belongs to the ATP-binding cassette (ABC) superfamily and consists of TAP1 (ABCB2) and TAP2 (ABCB3), each of which possesses a transmembrane and a nucleotide-binding domain (NBD). Here we describe the generation of recombinant Fv and Fab antibody fragments to human TAP from a hybridoma cell line expressing the TAP1-specific monoclonal antibody mAb148.3. The epitope of the antibody was mapped to the very last five C-terminal amino acid residues of TAP1 on solid-supported peptide arrays. The recombinant antibody fragments were heterologously expressed in Escherichia coli and purified to homogeneity from periplasmic extracts by affinity chromatography. The monoclonal and recombinant antibodies bind with nanomolar affinity to the last five C-terminal amino acid residues of TAP1 as demonstrated by ELISA and surface plasmon resonance. Strikingly, the recombinant antibody fragments confer thermal stability to the heterodimeric TAP complex. At the same time TAP is arrested in a peptide transport incompetent conformation, although ATP and peptide binding to TAP are not affected. Based on our results we suggest that the C terminus of TAP1 modulates TAP function presumably as part of the dimer interface of the NBDs.     

Quantitative approaches to computational vaccinology

This article reviews the newly released JenPep database and two new powerful techniques for T-cell epitope prediction: (i) the additive method; and (ii) a 3D-Quantitative Structure Activity Relationships (3D-QSAR) method, based on Comparative Molecular Similarity Indices Analysis (CoMSIA). The JenPep database is a family of relational databases supporting the growing need of immunoinformaticians for quantitative data on peptide binding to major histocompatibility complexes and to the Transporters associated with Antigen Processing (TAP). It also contains an annotated list of T-cell epitopes. The database is available free via the Internet (http://www.jenner.ac.uk/JenPep). The additive prediction method is based on the assumption that the binding affinity of a peptide depends on the contributions from each amino acid as well as on the interactions between the adjacent and every second side-chain. In the 3D-QSAR approach, the influence of five physicochemical properties (steric bulk, electrostatic potential, local hydrophobicity, hydrogen-bond donor and hydrogen-bond acceptor abilities) on the affinity of peptides binding to MHC molecules were considered. Both methods were exemplified through their application to the well-studied problem of peptides binding to the human class I MHC molecule HLA-A*0201.     

Tapasin interacts with the membrane-spanning domains of both TAP subunits and enhances the structural stability of TAP1 x TAP2 Complexes

The transporter associated with antigen processing (TAP) proteins are involved in transport of peptides from the cytosol into the endoplasmic reticulum. Two subunits, TAP1 and TAP2, are necessary and sufficient for peptide binding and peptide translocation across the endoplasmic reticulum membrane. TAP1 and TAP2 contain an N-terminal hydrophobic membrane-spanning region and a C-terminal nucleotide binding domain. Tapasin is an endoplasmic reticulum resident protein that has been found associated with the TAP subunits and shown to increase expression levels of TAP. Here we investigated TAP-tapasin interactions and their effects on TAP function in insect cells. We show tapasin binding to both TAP1 and TAP2 and to the corresponding nucleotide binding domain-exchanged chimeras as well as to a truncated TAP1.TAP2 complex containing just the membrane-spanning regions of TAP1 and TAP2. However, tapasin interactions with either the truncated TAP construct containing just the nucleotide binding domain are not observed. Tapasin is not required for high affinity peptide binding to TAP1.TAP2 complexes, and in fact, the presence of tapasin slightly reduces the affinity of TAP complexes for peptides. However, at near physiological temperatures, both tapasin and nucleotides stabilize the peptide binding site of TAP1.TAP2 complexes against inactivation, and enhanced thermostability of both TAP subunits is observed in the presence of tapasin. The enhanced structural stability of TAP1.TAP2 complexes in the presence of tapasin might explain the observations that tapasin increases TAP protein expression levels in mammalian cells.     

Accurate approximation method for prediction of class I MHC affinities for peptides of length 8, 10 and 11 using prediction tools trained on 9mers

Several accurate prediction systems have been developed for prediction of class I major histocompatibility complex (MHC):peptide binding. Most of these are trained on binding affinity data of primarily 9mer peptides. Here, we show how prediction methods trained on 9mer data can be used for accurate binding affinity prediction of peptides of length 8, 10 and 11. The method gives the opportunity to predict peptides with a different length than nine for MHC alleles where no such peptides have been measured. As validation, the performance of this approach is compared to predictors trained on peptides of the peptide length in question. In this validation, the approximation method has an accuracy that is comparable to or better than methods trained on a peptide length identical to the predicted peptides. AVAILABILITY: The algorithm has been implemented in the web-accessible servers NetMHC-3.0: http://www.cbs.dtu.dk/services/NetMHC-3.0, and NetMHCpan-1.1: http://www.cbs.dtu.dk/services/NetMHCpan-1.1