The Immunome Browser is a new feature integrated into the IEDB to more efficiently display large sets of immune epitope data. The Immunome Browser ‘maps’ query results from the IEDB onto a reference proteome. A reference proteome is used because the positions of epitopes and names of antigens reported by different investigators are inconsistent due to many versions of protein sequences and differences between strains. The Immunome Browser presents the mapped data along with calculated ‘epitope prominence scores’ that rank known epitopes based on how frequently subjects responded to each epitope, allowing identification of immunodominance.
Currently, only linear peptide epitopes with T cell or B cell responses are supported for viewing using the Immunome Browser. These limitations will be addressed in future versions.
To start using the Immunome Browser, just make a query as before and look for the new 'View In Immunome Browser' buttons, as shown below (Figure 1). Clicking on these buttons passes your query results to the Immunome Browser.
Figure 1. Search result summary page returned by the IEDB website.
Once in the Immunome Browser, the default parameters are set and the best matching reference proteome is displayed (Figure 2).
Figure 2. Default parameter setting for mapping query results to a selected reference proteome.
To override a default proteome or change other default mapping parameters, click ‘Change Settings’ to reveal additional fields.
After clicking Change Settings, all parameters are shown in Figure 3:
Figure 3. Parameters needed to map query results onto the reference proteome.
Use each of the fields to adjust mapping settings as follows:
This name refers to the best matching reference proteome for the query entered by the user. In the case where no match is found, the user must provide a set of antigens against which to map peptides. To provide the proteome or override a default proteome , use the 'File Proteome' field.
This button allows the user to submit a set of protein sequences, overriding the selected reference proteome.
Use 'Choose File' to provide your own antigen sequences in a FASTA format.
Peptide Similarity Threshold
This allows the user to change the similarity threshold between the epitope sequences and the protein they are being mapped to. A lower threshold may be desirable for proteins that display high sequence variability such as HCV.The sequence alignment step uses the USEARCH [http://www.drive5.com/usearch/] algorithm, to achieve fast sequence searches. Specifically, the command 'usearch search_local' is used. In our experience, usearch is faster than BLAST. Furthermore, USEARCH is a heuristic algorithm (BLAST is also heuristic). Hence, for low similarity thresholds, some valid epitopes may be missed.
This field allows the user to place the additional constraint that the proteins to which the epitopes are mapped must be homologous to the source protein of the epitopes, as described in the manuscript the epitope was described in.
Using this field the user decides between two different types of mappings; either ‘All’, where all possible epitope:protein matches are presented, or ‘Best’ where only the best matches are presented. In the case of 'Best' matches, ties are kept if multiple best matches are available.
Once all parameters are set as desired, click the ‘Submit’ button. This will initiate mapping of the query results to the chosen reference proteome.
Navigating the Mapping Results
After clicking the ‘Submit’ button, results of mapping query results to the reference proteome are displayed in Figure 4. Shown at the top is a navigation panel to display different types of mapping results. For example, by default, all proteins in the reference proteome are listed. If the user wants to see only those proteins with epitopes mapped, then ‘ORFs containing epitopes’ should be chosen.
In the bottom panel, either list of proteins or epitopes is shown based on the selection made in the navigation panel shown above. Of special note, in the case of ‘ORFs containing epitopes’, by clicking icons in column ‘Visualize’, one can visualize profiles of epitope prominence scores for corresponding antigens.
After clicking any one of the icons in column ‘Visualize’ from the page showing mapping results, a screen such as Figure 5 is shown. The top panel provides a visualization of where on the antigen epitopes are mapped. The x-axis shows residue positions and the y-axis shows epitope prominence scores. The bottom panel lists epitopes mapped to the antigen, their mapped positions and their epitope prominence scores (e.g. Response Freq. Score).
Different types of ‘epitope prominence scores’ are possible. By default, response frequency scores is chosen, which correlate with how often subjects responded to a given epitope. In addition, a score based on number of assays is also provided. As the name indicates, for each epitope, number of assays associated with it are collected and plotted. Thus, an epitope with higher ‘num assay’ score would mean that it has been studied more often. Additional details of the different epitope prominence scores are provided in the section that follows.
To illustrate usefulness of epitope prominence scores, a region in the HCV polyprotein containing the immunodominant epitope (NS3, 1248-1261)  is shown in Figure 6. In the figure, a peak in the region (1248-1261), which contains the immunodominant epitope, indicating relatively high response frequency scores is visible. In Figure 7, a list of peptides mapped to the region and sorted based on positions of their first residue is shown. Epitopes responsible for the peak visualized can be found based on response frequency scores listed in the last column.
Figure 4. Mapping results.
Figure 5. Visualization of mapped query results to an HCV polyprotein.
Figure 6. A zoom-in of a region of the HCV polyprotein containing the immunodominant epitope at positions 1248-1261 in the NS3 protein.
Figure 7. A screen shot of the table of peptides for the region containing the immunodominant epitope at positions 1248-1261 of the NS3 protein.
Different Types of Epitope Prominence Scores
As mentioned earlier, two different types of epitope prominence scores have been implemented, reflecting different ways an epitope's importance can be measured. Details of their calculations are provided below.
Response Frequency Scores
For a given peptide, its response frequency score is defined as: (responded – sqrt(responded))/subjects. The variables ‘responded’ and ‘subjects’ refer to number of subjects that gave positive immune response and total number of subjects, respectively. The term ‘sqrt(responded)’ is used to take into account statistical uncertainty associated with low counts. For example, response frequencies of 10/10 and 100/100 have the same ratios, however the latter has greater significance. Response frequency scores (RFScores) takes this into account, and thus the ratios have RFScores of 0.68 and 0.9, respectively.
Number of Assays
For a given peptide, num_assay scores is a count of its assay records. Thus those peptides that are studied more often will get higher num_assay scores. In addition, scores can be broken down into those assay records with positive or negative results.
1. Vita R, Zarebski L, Greenbaum JA, Emami H, Hoof I, Salimi N, Damle R, Sette A, Peters B. 2010. Nucleic Acids Res. The immune epitope database 2.0.
2. Kim Y, Vaughan K, Greenbaum J, Peters B, Law M, Sette A. 2012. PLoS One. A meta-analysis of the existing knowledge of immunoreactivity against hepatitis C virus (HCV).
3. Peptide:Protein search algorithm used: USEARCH (Version 6.0)
Parameters: “./usearch6.0.152_i86linux32 -search_local [fname_peptides] -db [fname_proteins] -evalue 10.0 -query_cov 1.0 -lopen 100.0 -lext 1.0 -blast6out [fname_output]”
4. Diepolder HM, Gerlach JT, Zachoval R, Hoffmann RM, Jung MC, et al. (1997) Immunodominant CD4+ T-cell epitope within nonstructural protein 3 in acute hepatitis C virus infection. J Virol 71: 6011–6019.