Shared Epitopes Scanner / Mapper

Scan an antigen/protein sequence for shared 8-mers and shared 11-mers against the selected proteome, and distinguish recognized but not shared sequences from unknown ones.

Input

Proteome:

Tip: pasted sequences can include spaces/newlines and will be cleaned automatically. Uploaded files should be plain-text FASTA (not gzip-compressed).

Results

Sequences
0
processed
Shared hits
0
unique k‑mers
Recognized
—
exact proteome match
Unknown
—
absent from proteome
Questions or problems? Please contact olivier.croce@univ-cotedazur.fr or patrick.brest@univ-cotedazur.fr.

How this tool works

This tool performs an exact identity screen. It does not predict real MHC binding, antigen processing, or immunogenicity. Instead, it helps you identify whether your sequence contains short peptide fragments that are also found in the selected proteome.

  1. Your input sequence is cleaned (letters only, converted to uppercase).
  2. The app extracts all unique 8-mers and 11-mers.
  3. These k-mers are compared with precomputed databases: data/<species>_8mers.json and data/<species>_11mers.json.
  4. If no shared k-mers are found, the app also checks whether the full cleaned sequence is present in the selected proteome using data/<species>_seqindex_sha256.json.

How to interpret the results

The per-sequence details panel can be used as a sanity check: it shows the cleaned sequence, the number of extracted k-mers, and offers an Export query k-mers (CSV) button with positions and sharing status.

Important: a shared 8-mer or 11-mer indicates exact sequence identity over that short window only. It should be interpreted as a screening signal, not as proof of cross-reactivity.

Browser note: the app loads its JSON databases from the local data/ folder using relative paths. This works on a normal web server and on http://localhost. Some browsers block these requests when the page is opened directly from disk (file://).

For reliable local testing, start a small web server in the folder containing the HTML file, for example: python3 -m http.server 8000 then open http://localhost:8000/.

Questions or problems? Contact olivier.croce@univ-cotedazur.fr or patrick.brest@univ-cotedazur.fr.