Medigene AG presented the web tool Expitope 3.0 at the 20th Annual Meeting of the Association for Cancer Immunotherapy (CIMT), taking place May 3-5, 2023, in Mainz, Germany. The Expitope web tool originates from a long standing collaboration between Medigene Immunotherapies GmbH and Prof.mitrij Frishman of the Faculty of Bioinformatics and his research team at the Technical University Munich. Through this joint endeavor, deep academic expertise in sophisticated bioinformatics, is applied for in silico assessment of potential target antigens for T cell receptor engineered T cell (TCR-T) therapies for treatment of solid cancer.

The web tool now coupled with artificial intelligence, draws from the vast and continually expanding data on the human genome and proteome of diverse tumors and healthy tissues. This web tool enables Medigene to employ cutting edge in silico technologies to efficiently investigate selected proteins and associated peptide epitopes as suitable targets for use in TCR-T therapies for selected cancer indications. The poster and oral presentation entitled, "Expitope 3.0 - An Advanced in silico Web tool Empowered with Machine Learning for Enhanced pHLA Epitope Prediction and Safety Assessment" will showcase the latest, publicly available version of Expitope 3.0, a faster and fully revised web tool, which will be launched in May/June 2023, to identify the target peptide HLA (pHLA) epitopes in antigens.

Comparing pHLA epitope expression in various healthy tissues allows to predict potential cross-reactivity and off-target toxicity thereby minimizing safety risks. Expitope 3.0 users can identify pHLA epitopes that could be suitable targets for TCR isolation and prediction of pHLA binding affinities as well as related epitopes with up to 50% mismatch for safety evaluation through the assessment of their expression patterns in healthy tissues. Expitope 3. 0 comes with an updated database, screening additionally for unique sequences on peptidome/transcriptome level thereby enlarging the number of screened epitopes.

The webtool utilizes machine learning to improve pHLA epitope binding prediction. An improved epitope prediction function provides higher accuracy in determining proteasomal/immunoproteasomal cleavage and presentation of epitopes. In summary, Expitope 3.0 allows for a quicker analysis with a broader search of databases and higher prediction accuracy of pHLA epitopes in antigens.

Its improved features and capabilities make it a valuable tool for researchers and clinicians working to identify potential targets for T cell therapies while minimizing safety risks. Expitope 3.0 is now available through the company's joint collaboration with Prof. Frishman and his research team to enable the company's to employ this faster and fully revised webtool encompassing information in greatly expanded owing the company's not only to screen for unique sequences as potential targets for T cell therapies but also to predict characteristics of peptide interactions with HLA molecules and their safety profiles in healthy tissues. Use of this advanced tool increases understanding of the safety profile needed for generation of highly differentiated TCR-T therapies that can address the unmet needs of patients with diverse solid tumors.