Identifying and validating new targets in order to overcome resistance to oncology treatments
Artificial intelligence (AI) is not a goal in itself, but rather a powerful technology to make Drug Discovery faster and more reliable.
The OncoSNIPER technology module is a pivotal manifestation of the key scientific competence of our Study activity. This module was designed to identify signatures that can stratify populations of patients refractory to cancer treatments. These signatures are then translated into therapeutic targets.
The OncoSNIPER platform integrates public, private, and proprietary data sources (from projects such as IMODI, OncoSNIPE®, etc.) and uses various artificial intelligence technologies (machine learning, deep learning, computer vision, natural language processing, etc.). One of the algorithms of OncoSNIPER is able to dig into the knowledge of our Drug Discovery experts, enabling a hybrid AI approach that combines the advantages of approaches based purely on data and expert system-type approaches.
OncoSNIPER also benefits from Oncodesign Services’ experimental capabilities, allowing us to generate ad hoc data and validate any results identified in silico at the preclinical stage. OPM thus identifies and selects targets based on an in silico scoring process and experimentation.
OncoSNIPER allows us to develop partnerships to discover new targets and biomarkers and to conclude licensing agreements on pre-identified therapeutic targets with pharmaceutical and biotechnological companies.
Our powerful databases: IMODI® and OncoSNIPE®
IMODI®: the missing link between cells and patients
IMODI® is a precision medicine unit that currently brings together scientific experts from 18 public and private organisations, with an initial plan for eight years and a private and public investment of €41 million.
Designed and coordinated by Oncodesign as part of a PIA2 PSPC [a French Government granted R&D program], the IMODI® unit has contributed to our four lines of research:
- Developing new experimental cancer models based on fully characterised, complete tumour specimens (data on the patient, biology, genomics, pharmacology, biobanking, etc.) xenografted on mice and immunodeficient rats.
- Modelling human tumour microenvironments in humanised transgenic mice
- Demonstrating the predictability of models using bioinformatics tools (databases and data mining)
- Studying the role of the intestinal microbiota in cancer
IMODI® capitalizes on medical and research skills around 10 key cancers: prostate, colon, breast, pancreatic, ovarian, lung, and liver cancers, as well as lymphoma, acute myeloid leukemia, and myeloma.
OncoSNIPE®: guide your therapeutic solutions for patients refractory to cancer treatment
The OncoSNIPE® programme aims to stratify and characterise populations of patients refractory to cancer treatments.
Based on bioinformatics, artificial intelligence, statistical learning, and semantic enrichment approaches, OncoSNIPE® guides and predicts a patient’s potential to benefit from treatment, enables the discovery of new therapeutic targets, and reduces treatment failure rates.
OncoSNIPE® is a six-year PIA3 PSPC [a French Government granted R&D program] run in partnership with Expert AI, Coexia, Acobiom, 11 hospitals in France, and the Unicancer federation. The OncoSNIPE® database was generated using the results of a clinical trial (NCT04548960) designed, managed, and sponsored by OPM.
We collect clinical, biological, genetic, and imaging data and samples from more than 600 chemotherapy-naive patients with breast cancer, pancreatic cancer, and lung cancer during the course of their treatments, as well as several hundred others in retrospect.
We implement semantic enrichment methods based on medical records, the integration of a priori knowledge from biological networks, predictive modelling, and longitudinal modelling capabilities.
We are able to provide new, validated therapeutic targets to our precision medicine platform to support our own therapeutic research and that of our partners.