The Epitope Location Finder performs 3 analyses on a submitted peptide. It is an interface that is meant to be a workbench for experimentalists who have T-cell reactivity data for peptides and want to quickly define probable epitopes within their peptides based on HLA anchor motifs or to find information about previously described epitopes stored in the immunology database. It can also help identify potentially missed T-cell reactivities due to variations in the sequence strain selected as a basis for the peptides used to test the response. To get a feel for how the site works, run the "Sample Input".
Enter a polypeptide sequence of interest, such as an immunologically active peptide. Sequences must be:
ELF performs 3 analyses:
(1) Show known epitopes. When checked, ELF will find the coordinates of your query sequence and search for known epitopes in our HIV CTL and Helper epitope databases. It will list all epitopes within the coordinates of your query, regardless of HLA and regardless of sequence mismatch. If you choose any specific HLAs in the second analysis, these submitted HLAs will be flagged.
(2) Show potential epitopes based on anchor residues. This analysis searches your protein for anchor motifs of the chosen HLA type(s).
Choosing HLAs/MHCs for potential epitope analysis. There are two selection menus, one for genotypes and another for serotypes. Both Class I and Class II alleles are available. Any number of items can be chosen from each list; use a shift-click or control-click to make multiple selections. Note that the selection of genotypes is synchronized with the selection of genotypes in the "IEDB binding predictions" analysis, if selected.
(3) Show potential epitopes based on IEDB binding predictions. This analysis provides predictions for the strength of binding between peptides from your query and Class I and Class II alleles. Details about the algorithms and scores can be obtained from IEDB T Cell Epitope Prediction Tools.
Choosing HLAs for IEDB binding analysis. Note that, whenever possible, the selection of genotypes is automatically synchronized with the selection of genotypes (but not serotypes) in the "Show potential epitopes based on anchor residues" analysis.
The first part of the output, reproduced here, presents a set of links to your results. The output of each of these links is shown below.
View Genomic location of your peptide.
View All anchor motifs used in this analysis.
View Potential "epitopes" ordered by HLAs.
View Database records for known CTL epitopes overlapping your query, regardless of HLA.
View Database records for known Helper epitopes overlapping your query, regardless of HLA.
The first link, Genomic location of your peptide, connects to a page with summary information about the location of the submitted protein as mapped onto the reference sequence genome. The table gives location information in tabular form, and the alignment shows your sequence aligned to the standard HIV reference sequence.
Table of protein regions touched by query sequence. AA = amino acid, NA = nucleic acid.
CDS AA position relative to query sequence start AA position relative to polyprotein start in reference sequence AA position relative to protein start in reference sequence NA position relative to CDS start in reference sequence NA position relative to reference sequence genome start Name of Protein 1 -> 36 25 -> 60 25 -> 60 73 -> 180 414 -> 521
Alignment of the query sequence to the HIV reference sequence (Similarity % = 100.0)Query PGGGQIVGGV YLLPRRGPRL GVRATRKTSE RSQPRG 36 :::::::::: :::::::::: :::::::::: :::::: ReferenceSeq PGGGQIVGGV YLLPRRGPRL GVRATRKTSE RSQPRG
The Anchor Motifs link shows any known anchor residue motifs associated with the HLAs submitted. In the motifs column a "." character means any amino acid may occur at that position, while square brackets list the amino acids required at that position. For example, in the A*0201 motifs below, a L or an M must occur at position 2.
HLA anchor residue motifs used are taken from our Motif Scan tool. These motifs can be found in The HLA Fact Book by Marsh, Parham, and Barber; MHC Ligands and Peptide Motifs, by Rammensee, Bachmann, and Stevanovic; the SYFPEITHI web site; and other published sources. For more detailed information concerning HLA motifs, see Motif Scan.
Anchor Residue Motifs
The Potential Epitopes table, shows that at position 12-20 of the submitted sequence "PGGGQIVGGVYLLPRRGPRLGVRATRKTSERSQPRG" there occurs an amino acid string which matches one of the A*0201 anchor residue motifs in the Anchor Motifs Table above.
pgggqivggvyLLPRRGPRLgvratrktsersqprg submitted peptide ||||||||| matches .L......L a A*0201 anchor motif
Note that these potential "epitopes" are purely theoretical; it is possible that they have never been observed, and do not necessarily correspond to any known epitopes. In the table they are ordered by HLA.
Position in query peptide AA sequence HLA Anchor motif (12-20) LLPRRGPRL A*0201 .[LM]......[VL] (11-20) YLLPRRGPRL A*0201 .[LM].......[VL] (12-20) LLPRRGPRL A*0202 .[L]......[LV] (11-20) YLLPRRGPRL A*0202 .[L].......[LV] (12-20) LLPRRGPRL A*0204 .[L]......[L] etc. etc. etc. etc.
The last links, Database Records, present CTL and Helper search results. Database records for all known epitopes overlapping the region of the query sequence are listed, regardless of HLA and regardless of sequence mismatch.
These results (except the ones that overlap the ends of your query) are also shown graphically as an alignment in the next section of results; see next section below.
This part of the ELF output displays all epitopes from the HIV immunology database within the bounds of your submitted protein sequence. Click the epitope itself to display its full database record. The HLA, if defined, is listed after each epitope. The button, when clicked, will use QuickAlign to align that epitope to all sequences in our most recent "Complete genome alignment".
Letters in red are residues that differ from your query sequence. These residues may have bearing on the HLA binding of the corresponding peptide in your query. The symbol means the HLA of the epitope matches one of your selected HLAs.
These peptides have anchor residues that match one or more motifs associated with the submitted HLA. The alignment below presents them in genomic order. They are also available in HLA order in the clickable links earlier in this output.
Note that these potential "epitopes" are purely theoretical; it is possible that they have never been observed, and do not necessarily correspond to any known epitopes.
PGGGQIVGGVYLLPRRGPRLGVRATRKTSERSQPRG GQIVGGVYL (A*0205 .[VLIMQ]......[L]) GQIVGGVYL (A*0214 .[VQL]......[LV]) GQIVGGVYLL (A*0205 .[VLIMQ].......[L]) GQIVGGVYLL (A*0214 .[VQL].......[LV]) QIVGGVYL (A*0205 .[VLIMQ].....[L]) QIVGGVYLL (A*0205 .[VLIMQ]......[L]) IVGGVYLL (A*0205 .[VLIMQ].....[L]) IVGGVYLL (A*0214 .[VQL].....[LV]) LPRRGPRL (B7 .[P].....[LF]) LPRRGPRL (B*0702 .[P].....[L]) LPRRGPRL (B*0703 .[P].....[L]) LPRRGPRL (B*0705 .[P].....[L])
This analysis runs the Class I and Class II Binding Predictions tools at the Immune Epitope Database (IEDB), http://www.immuneepitope.org/.
Choose the MHC alleles of interest for class I and class II predictions. Choose to display the single best binder, or all binders below a percentile cutoff. Note that, whenever possible, the selection of alleles is automatically synchronized with the selection of genotypes (but not serotypes) in the "Show potential epitopes" analysis.
Percentile. Results are grouped by percentile, with low percentiles indicating better binding. Note: sequences with a low percentile may still be very unlikely to bind! To truly assess the likelihood of binding, you will need closely examine the output on the IEDB page!
From the ELF output page, click the allele name to view the IEDB results. IMPORTANT: click the "expand the result" box to view the full IEDB results. The IEDB results will include IC50 values from multiple prediction methods. As a rough guideline, peptides with IC50 values <50 nM are considered high affinity, 50-500 nM intermediate affinity and >5000 nM low affinity. No known T-cell epitope has an IC50 value greater than 5000.
Additional information for interpreting the IEDB results can be obtained from: