Epibase^® is our structure based approach for in silico T-cell epitope identification. The platform searches for potential epitopes in complex with HLA receptors in order to determine their affinity.

T-cell epitope identification algorithms have been around for over 15 years. Previous generation methods are mainly driven by the (sequence) analysis of observed T-cell epitopes.
While these methodologies may work for HLA receptors where a lot of experimentally described epitopes are known, they fail to identify epitopes for those HLA alleles where less data is available.
Because of its structure based approach, Epibase^® can identify epitopes for any HLA sub-type, including those for which little or no experimental data is available.

Figure: Epibase^® assesses the immunogenicity of proteins and protein regions.

Blue: Epibase^® results
Black: Observed T-cell response in 50 healthy donors
(J. Immunol., 2002, 168: 155-161)

HIGH POPULATION COVERAGE
Epibase^® addresses HLA Class-I and Class-II allotypes which are present in at least 3% of given populations (such as Caucasian, Oriental, etc,...). Any allotype can be added upon request.

BUILD NEW INTELLECTUAL PROPERTY
Epibase^® has the intrinsic capacity to identify epitopes with atypical sequence motifs. Often, such epitopes are not recognized by learning based methods.

HIGH ACCURACY
Epibase^® is significantly more accurate in epitope identification than sequence based methodologies.
Studying the interaction between peptides and their receptors translates into a larger information content than merely using the sequences of the peptides.

FAST SCREENING
Epibase^® provides low cost information compared to experimental assays and ensures fast screening compared to peptide synthesis and MHC presentation experiments.

STANDARDIZED PROFILES
Epibase^® is being used by a wide variety of life sciences companies to compare and address immunogenicity.
Therefore the technology allows easy comparison of data throughout proteins and protein leads.

Collaborations

Figure: Epibase^® shows increased accuracy compared to previous generation methods. (Observed (X), Predicted (Y))