Description du projet
Un nouveau logiciel pour concevoir des vaccins contre le cancer spécifiques aux patients
Des scientifiques espèrent qu’une immunothérapie personnalisée pourrait guérir le cancer, mais cela implique une vaccination personnalisée, soit un processus de laboratoire complexe, chronophage et coûteux pour cibler un cas spécifique de mutation. Pour cette raison, des chercheurs développent des algorithmes pour identifier les néoantigènes immunogènes par le biais des données du séquençage nouvelle génération (NGS) obtenues à partir d’échantillons de la tumeur. Le projet MEDIVAC, financé par l’UE, a pour objectif de soutenir ce processus en appliquant le cadre d’apprentissage machine d’OncoImmunity (OI). Ce dernier permet d’utiliser plusieurs bases de données, notamment publiques, pour accroître de manière significative la performance du processus d’identification des néoantigènes pertinents. Cela rapprochera grandement la science médicale de l’introduction d’une vaccination personnalisée prometteuse pour le traitement anticancéreux.
Objectif
Cancer is arguably the most feared of all diseases, destroying lives regardless of the age of its victims. Immunotherapies are currently regarded as the most promising avenue to delivering the holy grail of medicine i.e. providing a cure for cancer. Despite their Nobel-winning status, personalisation of immunotherapies remains akey challenge to which no cost-effective solution currently exists. Current methods for identifying the immunogenic neoantigens required to design patient-specific cancer vaccines typically utilize next generation sequencing (NGS) analysis of DNA and RNA coupled with wet lab methods (e.g. spectroscopy). However, these approaches are time consuming to perform, expensive and not readily scalable–which currently prohibits the mass roll-out of personalised cancer vaccines.
Despite the fact that intensive research has been dedicated to developing prediction algorithms which can identify immunogenic neoantigens from NGS data from tumor samples, their accuracy has not yet reached a competitive performance compared to wet lab methods. To bridge this gap, OncoImmunity (OI) has developed a comprehensive machine learning framework, trained using public and proprietary datasets to optimise performance. Once fed with patient NGS data from healthy and tumor tissue, OI’s algorithms identifies the most clinically relevant neoantigen candidates, with an unmatched accuracy (4-fold increase of prediction accuracy), which can be subsequently engineering into personalised vaccine cancer constructs.
Considering the potential of personalised therapy within cancer immunotherapy, OI’s core technology meets all the requirements to become a key enabling technology, providing cost-effective, scalable and sensitive identification of clinically relevant targets for vaccine development. Thus, it has the potential to serve as a cornerstone to revolutionize the fight against cancer – while untapping an immense business opportunity to fuel our company’s growth.
Champ scientifique
- medical and health sciencesbasic medicinepharmacology and pharmacypharmaceutical drugsvaccines
- medical and health sciencesclinical medicineoncology
- natural sciencesbiological sciencesgeneticsRNA
- medical and health sciencesbasic medicineimmunologyimmunotherapy
- natural sciencescomputer and information sciencesartificial intelligencemachine learning
Programme(s)
Régime de financement
SME-2 - SME instrument phase 2Coordinateur
0379 OSLO
Norvège
L’entreprise s’est définie comme une PME (petite et moyenne entreprise) au moment de la signature de la convention de subvention.