Sanne Lise Lauemøller

Reliable prediction of T-cell epitopes using neural networks with novel sequence representations

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.

Tumor‐associated antigens identified by mRNA expression profiling induce protective anti‐tumor immunity

Defined tumor‐associated antigens (TAA) are attractive targets for anti‐tumor immunotherapy. Here, we describe a novel genome‐wide approach to identify multiple TAA from any given tumor. A panel of transplantable thymomas was established from an inbred p53–/– mouse strain. The resulting tumors were examined for gene expression by mRNA microarray scanning. This analysis revealed heterogeneity of the tumors in agreement with the assumption that they represent different tumorigenic events. Several genes were overexpressed in one or more of the tumors. To examine whether overexpressed genes might be used to identify TAA, mice were immunized with mixtures of peptides representing putative cytotoxic T cell epitopes derived from one of the gene products. Indeed, such immunized mice were partially protected against subsequent tumor challenge. Despite being immunized with bona fide self antigens, no clinical signs of autoimmune reactions were observed. Thus, it appears possible to evaluate the entire metabolism of any given tumor and use this information rationally to identify multiple epitopes of value in the generation of tumor‐specific immunotherapy. We expect that human tumors express similar tumor‐specific metabolic imprints, which may be used to identify patient‐specific arrays of TAA. This may enable a multi‐epitope based immunotherapy with improved prospects of clinical tumor rejection.

Selecting informative data for developing peptide-MHC binding predictors using a query by committee approach

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.

Measurement of Peptide‐MHC Interactions in Solution Using the Spin Column Filtration Assay

This unit describes how peptide‐MHC complexes can be generated in vitro using affinity‐purified MHC and synthetic peptide. The unit first describes how the interaction between peptide and MHC interaction can be measured in an accurate, quantitative biochemical assay. This procedure has been optimized for efficient separation of free peptide and MHC‐bound peptide through a novel principle, termed “gradient centrifugation.” The first two support protocols describe how to set up a biochemical fluid‐phase binding reaction between peptide and MHC class I and class II, respectively. Also, an alternative procedure for setting up a biochemical fluid phase binding reaction between β2m and MHC class I is included. Finally a more versatile inhibition assay is described. The assay is simple and robust, and has several advantages compared to the classical gel‐filtration assay, including increased sensitivity and throughput. It also demands fewer resources both in terms of unique reagents and labor, and it generates less hazardous waste. Thus, the spin column gel‐filtration assay is ideal for routine work.