Search

Recognition of peptides bound to major histocompatibility complex (MHC) class I molecules by T lymphocytes is an essential part of immune surveillance. Each MHC allele has a characteristic peptide binding preference, which can be captured in prediction algorithms, allowing for the rapid scan of entire pathogen proteomes for peptide likely to bind MHC. Here we make public a large set of 48,828 quantitative peptide-binding affinity measurements relating to 48 different mouse, human, macaque, and chimpanzee MHC class I alleles. We use this data to establish a set of benchmark predictions with one neural network method and two matrix-based prediction methods extensively utilized in our groups. In general, the neural network outperforms the matrix-based predictions mainly due to its ability to generalize even on a small amount of data. We also retrieved predictions from tools publicly available on the internet. While differences in the data used to generate these predictions hamper direct comparisons, we do conclude that tools based on combinatorial peptide libraries perform remarkably well. The transparent prediction evaluation on this dataset provides tool developers with a benchmark for comparison of newly developed prediction methods. In addition, to generate and evaluate our own prediction methods, we have established an easily extensible web-based prediction framework that allows automated side-by-side comparisons of prediction methods implemented by experts. This is an advance over the current practice of tool developers having to generate reference predictions themselves, which can lead to underestimating the performance of prediction methods they are not as familiar with as their own. The overall goal of this effort is to provide a transparent prediction evaluation allowing bioinformaticians to identify promising features of prediction methods and providing guidance to immunologists regarding the reliability of prediction tools.

Reverse immunogenetic approaches attempt to optimize the selection of candidate epitopes, and thus minimize the experimental effort needed to identify new epitopes. When predicting cytotoxic T cell epitopes, the main focus has been on the highly specific MHC class I binding event. Methods have also been developed for predicting the antigen-processing steps preceding MHC class I binding, including proteasomal cleavage and transporter associated with antigen processing (TAP) transport efficiency. Here, we use a dataset obtained from the SYFPEITHI database to show that a method integrating predictions of MHC class I binding affinity, TAP transport efficiency, and C-terminal proteasomal cleavage outperforms any of the individual methods. Using an independent evaluation dataset of HIV epitopes from the Los Alamos database, the validity of the integrated method is confirmed. The performance of the integrated method is found to be significantly higher than that of the two publicly available prediction methods BIMAS and SYFPEITHI. To identify 85% of the epitopes in the HIV dataset, 9% and 10% of all possible nonamers in the HIV proteins must be tested when using the BIMAS and SYFPEITHI methods, respectively, for the selection of candidate epitopes. This number is reduced to 7% when using the integrated method. In practical terms, this means that the experimental effort needed to identify an epitope in a hypothetical protein with 85% probability is reduced by 20-30% when using the integrated method. The method is available at http://www.cbs.dtu.dk/services/NetCTL. Supplementary material is available at http://www.cbs.dtu.dk/suppl/immunology/CTL.php.

Cytotoxic T cells (CTLs) perceive the world through small peptides that are eight to ten amino acids long. These peptides (epitopes) are initially generated by the proteasome, a multi-subunit protease that is responsible for the majority of intra-cellular protein degradation. The proteasome generates the exact C-terminal of CTL epitopes, and the N-terminal with a possible extension. CTL responses may diminish if the epitopes are destroyed by the proteasomes. Therefore, the prediction of the proteasome cleavage sites is important to identify potential immunogenic regions in the proteomes of pathogenic microorganisms (or humans). We have recently shown that NetChop, a neural network-based prediction method, is the best method available at the moment to do such predictions; however, its performance is still lower than desired. Here, we use novel sequence encoding methods and show that the new version of NetChop predicts approximately 10% more of the cleavage sites correctly while lowering the number of false positives with close to 15%. With this more reliable prediction tool, we study two important questions concerning the function of the proteasome. First, we estimate the N-terminal extension of epitopes after proteasomal cleavage and find that the average extension is relatively short. However, more than 30% of the peptides have N-terminal extensions of three amino acids or more, and thus, N-terminal trimming might play an important role in the presentation of a substantial fraction of the epitopes. Second, we show that good TAP ligands have an increased chance of being cleaved by the proteasome, i.e., the specificity of TAP has evolved to fit the specificity of the proteasome. This evolutionary relationship allows for a more efficient antigen presentation.

Motivation: Prediction of which peptides will bind a specific major histocompatibility complex (MHC) constitutes an important step in identifying potential T-cell epitopes suitable as vaccine candidates. MHC class II binding peptides have a broad length distribution complicating such predictions. Thus, identifying the correct alignment is a crucial part of identifying the core of an MHC class II binding motif. In this context, we wish to describe a novel Gibbs motif sampler method ideally suited for recognizing such weak sequence motifs. The method is based on the Gibbs sampling method, and it incorporates novel features optimized for the task of recognizing the binding motif of MHC classes I and II. The method locates the binding motif in a set of sequences and characterizes the motif in terms of a weight-matrix. Subsequently, the weight-matrix can be applied to identifying effectively potential MHC binding peptides and to guiding the process of rational vaccine design., Results: We apply the motif sampler method to the complex problem of MHC class II binding. The input to the method is amino acid peptide sequences extracted from the public databases of SYFPEITHI and MHCPEP and known to bind to the MHC class II complex HLA-DR4(B1*0401). Prior identification of information-rich (anchor) positions in the binding motif is shown to improve the predictive performance of the Gibbs sampler. Similarly, a consensus solution obtained from an ensemble average over suboptimal solutions is shown to outperform the use of a single optimal solution. In a large-scale benchmark calculation, the performance is quantified using relative operating characteristics curve (ROC) plots and we make a detailed comparison of the performance with that of both the TEPITOPE method and a weight-matrix derived using the conventional alignment algorithm of ClustalW. The calculation demonstrates that the predictive performance of the Gibbs sampler is higher than that of ClustalW and in most cases also higher than that of the TEPITOPE method.

Major histocompatibility complex (MHC) proteins are encoded by extremely polymorphic genes and play a crucial role in immunity. However, not all genetically different MHC molecules are functionally different. Sette and Sidney (1999) have defined nine HLA class I supertypes and showed that with only nine main functional binding specificities it is possible to cover the binding properties of almost all known HLA class I molecules. Here we present a comprehensive study of the functional relationship between all HLA molecules with known specificities in a uniform and automated way. We have developed a novel method for clustering sequence motifs. We construct hidden Markov models for HLA class I molecules using a Gibbs sampling procedure and use the similarities among these to define clusters of specificities. These clusters are extensions of the previously suggested ones. We suggest splitting some of the alleles in the A1 supertype into a new A26 supertype, and some of the alleles in the B27 supertype into a new B39 supertype. Furthermore the B8 alleles may define their own supertype. We also use the published specificities for a number of HLA-DR types to define clusters with similar specificities. We report that the previously observed specificities of these class II molecules can be clustered into nine classes, which only partly correspond to the serological classification. We show that classification of HLA molecules may be done in a uniform and automated way. The definition of clusters allows for selection of representative HLA molecules that can cover the HLA specificity space better. This makes it possible to target most of the known HLA alleles with known specificities using only a few peptides, and may be used in construction of vaccines. Supplementary material is available at http://www.cbs.dtu.dk/researchgroups/immunology/supertypes.html.

Strategies for selecting informative data points for training prediction algorithms are important, particularly when data points are difficult and costly to obtain. A Query by Committee (QBC) training strategy for selecting new data points uses the disagreement between a committee of different algorithms to suggest new data points, which most rationally complement existing data, that is, they are the most informative data points. In order to evaluate this QBC approach on a real-world problem, we compared strategies for selecting new data points. We trained neural network algorithms to obtain methods to predict the binding affinity of peptides binding to the MHC class I molecule, HLA-A2. We show that the QBC strategy leads to a higher performance than a baseline strategy where new data points are selected at random from a pool of available data. Most peptides bind HLA-A2 with a low affinity, and as expected using a strategy of selecting peptides that are predicted to have high binding affinities also lead to more accurate predictors than the base line strategy. The QBC value is shown to correlate with the measured binding affinity. This demonstrates that the different predictors can easily learn if a peptide will fail to bind, but often conflict in predicting if a peptide binds. Using a carefully constructed computational setup, we demonstrate that selecting peptides with a high QBC performs better than low QBC peptides independently from binding affinity. When predictors are trained on a very limited set of data they cannot be expected to disagree in a meaningful way and we find a data limit below which the QBC strategy fails. Finally, it should be noted that data selection strategies similar to those used here might be of use in other settings in which generation of more data is a costly process.

In this paper we describe an improved neural network method to predict T-cell class I epitopes. A novel input representation has been developed consisting of a combination of sparse encoding, Blosum encoding, and input derived from hidden Markov models. We demonstrate that the combination of several neural networks derived using different sequence-encoding schemes has a performance superior to neural networks derived using a single sequence-encoding scheme. The new method is shown to have a performance that is substantially higher than that of other methods. By use of mutual information calculations we show that peptides that bind to the HLA A*0204 complex display signal of higher order sequence correlations. Neural networks are ideally suited to integrate such higher order correlations when predicting the binding affinity. It is this feature combined with the use of several neural networks derived from different and novel sequence-encoding schemes and the ability of the neural network to be trained on data consisting of continuous binding affinities that gives the new method an improved performance. The difference in predictive performance between the neural network methods and that of the matrix-driven methods is found to be most significant for peptides that bind strongly to the HLA molecule, confirming that the signal of higher order sequence correlation is most strongly present in high-binding peptides. Finally, we use the method to predict T-cell epitopes for the genome of hepatitis C virus and discuss possible applications of the prediction method to guide the process of rational vaccine design.

Background: For most sequenced prokaryotic genomes, about a third of the protein coding genes annotated are "orphan proteins", that is, they lack homology to known proteins. These hypothetical genes are typically short and randomly scattered throughout the genome. This trend is seen for most of the bacterial and archaeal genomes published to date., Results: In contrast we have found that a large fraction of the genes coding for such orphan proteins in the Methanopyrus kandleri AV19 genome occur within two large regions. These genes have no known homologs except from other M. kandleri genes. However, analysis of their lengths, codon usage, and Ribosomal Binding Site (RBS) sequences shows that they are most likely true protein coding genes and not random open reading frames., Conclusions: Although these regions can be considered as candidates for massive lateral gene transfer, our bioinformatics analysis suggests that this is not the case. We predict many of the organism specific proteins to be transmembrane and belong to protein families that are non-randomly distributed between the regions. Consistent with this, we suggest that the two regions are most likely unrelated, and that they may be integrated plasmids.

Secondary structure prediction involving up to 800 neural network predictions has been developed, by use of novel methods such as output expansion and a unique balloting procedure. An overall performance of 77.2%-80.2% (77.9%-80.6% mean per-chain) for three-state (helix, strand, coil) prediction was obtained when evaluated on a commonly used set of 126 protein chains. The method uses profiles made by position-specific scoring matrices as input, while at the output level it predicts on three consecutive residues simultaneously. The predictions arise from tenfold, cross validated training and testing of 1032 protein sequences, using a scheme with primary structure neural networks followed by structure filtering neural networks. With respect to blind prediction, this work is preliminary and awaits evaluation by CASP4.

Phosphoribosyltransferases catalyze the formation of nucleotides from a nitrogenous base and 5-phosphoribosyl-alpha-1-pyrophosphate (PRPP). These enzymes and the PRPP synthases resemble each other in a short homologous sequence of 13 amino acid residues which has been termed the PRPP binding site and which interacts with the ribose 5-phosphate moiety in structurally characterized complexes of PRPP and nucleotides. We show that each class of phosphoribosyltransferases has subtle deviations from the general consensus PRPP binding site and that all uracil phosphoribosyltransferases (UPRTases) have a proline residue at a position where other phosphoribosyltransferases and the PRPP synthases have aspartate. To investigate the role of this unusual proline (Pro 131 in the E. coli UPRTase) for enzyme activity, we changed the residue to an aspartate and purified the mutant P131D enzyme to compare its catalytic properties with the properties of the wild-type protein. We found that UPRTase of E. coli obeyed the kinetics of a sequential mechanism with the binding of PRPP preceding the binding of uracil. The basic kinetic constants were derived from initial velocity measurements, product inhibition, and ligand binding assays. The change of Pro 131 to Asp caused a 50-60-fold reduction of the catalytic rate (kcat) in both directions of the reaction and approximately a 100-fold increase in the KM for uracil. The KM for PRPP was strongly diminished by the mutation, but kcat/KM,PRPP and the dissociation constant (KD,PRPP) were nearly unaffected. We conclude that the proline in the PRPP binding site of UPRTase is of only little importance for binding of PRPP to the free enzyme, but is critical for binding of uracil to the enzyme-PRPP complex and for the catalytic rate.

Orotate phosphoribosyltransferase (OMP synthase, EC 2.4.2.10) forms the UMP precursor orotidine 5'-monophophate (OMP) from orotate and alpha-D-5-phosphoribosyl-1-pyrophosphate (PRPP). Here, equilibrium binding, isotope partitioning, and chemical quench studies were used to determine rate and equilibrium constants for the kinetic mechanism. PRPP bound to two sites per dimer with a KD of 33 microM. Binding of OMP and orotate also occurred to a single class of two sites per dimer, with KD values of 3 and 280 microM, respectively. Pyrophosphate binding to two sites was weak with a KD of 960 microM, and in the presence of bound orotate, its affinity for the first site was enhanced 4-fold (KD = 230 microM). Preformed E.OMP, E.PRPP, E.PPi, and E.orotate complexes were trapped as products in isotope partitioning experiments, indicating that each was catalytically competent and confirming a random mechanism. Rapid quench experiments revealed burst kinetics for product formation in both the forward phosphoribosyltransferase and the reverse pyrophosphorolysis reactions. The steady-state rate in the forward reaction was preceded by a burst (nfwd = 1.5/dimer) of at least 300 s-1. In the pyrophosphorolysis reaction, a burst (nrev = 0.7/dimer; k >/= 300 s-1) was also noted. These results allowed us to develop a complete kinetic mechanism for OPRTase, in which a rapid phosphoribosyl transfer reaction at equilibrium is followed by a slow step involving release of product. When the microviscosity, etarel, of the reaction medium was increased with sucrose, the forward kcat decreased in proportion to etarel with a slope of 0.8. In the reverse reaction a more limited dependence of kcat (slope = 0. 3) was observed. On the basis of the known structures of OPRTase, we propose that a highly conserved, catalytically important, solvent-exposed loop descends during catalysis to shield the active site. In the accompanying paper, the slow product release step is shown to relate to movement of the solvent-exposed loop.