Periodic Reporting for period 1 - mCUBE (Multi-omic data driven drug discovery and indication prioritisation platform in oncology)
Période du rapport: 2022-08-01 au 2023-04-30
The attrition rate in oncology drug discovery is among the highest, causing a real societal challenge. Most cancers remain unmet clinical needs and only 1 in 5 patients respond to their treatment. Drug development takes an average of 10 to 15 years and requires more than €2bn in investment to bring an efficient compound from the bench to the bedside. In 2020, 80% of clinical trials were focusing on cancer and a third of them were testing anticancer drugs. Accumulating evidence suggests that drug candidates validated through molecular analysis (“omics” analysis) are twice as likely to succeed in clinical studies. However, with data growing by about 25 petabytes year on year and the lack of adequate tools for data aggregation and analysis, cancer scientists battle to leverage this rich data for the prioritisation of drug discovery programs. They need an efficient and user-friendly solution to support their daily work.
The first part of this project has focused on further developing mCUBE to include more data sources, more indications, more omics, and more analysis pipelines.
More specifically, our team carried out the following activities:
- Development of AI-augmented pipelines for the identification and integration of datasets,
- Development of multi-omic data analysis pipelines for target discovery, biomarker discovery, patient selection and preclinical model selection,
- Implementation of a subset of these tools into a user-friendly cloud-based platform and deployment of the solution for our industrial partners, and
- Generation of scientific results using the different tools and pipelines developed with our academic partners.
Our main achievements relate to the outcomes of our partnerships. Our platform continues to receive outstanding feedback from our biopharma customers, with users volunteering to become mCUBE ambassadors within their organisation. After only a few months of deployment at one of our industrial partners, mCUBE has not only become one of the bioinformatic tools they use on a daily basis, it has also been utilised to help design clinical trials. Our team also generated their first precision oncology insights with academic partners. More specifically, in the context of a partnership with the Eyquem’s lab at UCSF, a list of promising targets has been prioritised and is now being tested experimentally for the treatment of multiple myeloma, colorectal cancer and gastric cancer. The first part of this work has already been presented in a poster entitled “Integrating transcriptomics and proteomics for the discovery of novel antigen targets on the surface of malignant plasma cells amenable to CAR-T cell approach in the treatment of RRMM patients” at the 2022 American Society of Hematology (ASH) Annual Meeting.
More specifically, our team carried out the following activities:
- Development of AI-augmented pipelines for the identification and integration of datasets,
- Development of multi-omic data analysis pipelines for target discovery, biomarker discovery, patient selection and preclinical model selection,
- Implementation of a subset of these tools into a user-friendly cloud-based platform and deployment of the solution for our industrial partners, and
- Generation of scientific results using the different tools and pipelines developed with our academic partners.
Our main achievements relate to the outcomes of our partnerships. Our platform continues to receive outstanding feedback from our biopharma customers, with users volunteering to become mCUBE ambassadors within their organisation. After only a few months of deployment at one of our industrial partners, mCUBE has not only become one of the bioinformatic tools they use on a daily basis, it has also been utilised to help design clinical trials. Our team also generated their first precision oncology insights with academic partners. More specifically, in the context of a partnership with the Eyquem’s lab at UCSF, a list of promising targets has been prioritised and is now being tested experimentally for the treatment of multiple myeloma, colorectal cancer and gastric cancer. The first part of this work has already been presented in a poster entitled “Integrating transcriptomics and proteomics for the discovery of novel antigen targets on the surface of malignant plasma cells amenable to CAR-T cell approach in the treatment of RRMM patients” at the 2022 American Society of Hematology (ASH) Annual Meeting.
No performant multi-omic data aggregation and analysis tool is currently available on the market and cancer scientists are thus unable to leverage this valuable data for the prioritisation of drug discovery programs. With limited data skills, they can only rely on often disappointing collaborations with teams of bioinformaticians, with whom communication can be difficult, or on the use of suboptimal data science software. While the available solutions are generally based on relevant mathematical frameworks, they are poorly designed, non-user friendly and require computer programming skills. Moreover, available tools do not use AI to assist the user in the harmonisation of clinical data, a very time-consuming step in data aggregation.
We are developing mCUBE, an Artificial Intelligence (AI)-augmented cloud-based precision oncology platform for multi-omic, multi-cancer and multi-source data identification, aggregation and analysis. mCUBE encapsulates Epigene Labs’ deep expertise in oncology, data science, and bioinformatics into a user-friendly solution that will make a difference for cancer patients. Our solution delivers 3 main gains:
- increased productivity for our users (cancer scientists),
- increased profitability for our customers (biopharma companies), and
- improved care for cancer patients ultimately.
The conventional timeline for discovering a new target for a specific indication is 2 years. At Epigene, in the context of our first partnership with the Eyquem’s lab at UCSF, we not only rediscovered known antigens but also prioritised a list of promising targets for relapsed/refractory multiple myeloma in only 6 months. A few months later, in a second program focusing on colorectal and gastric cancer, this same task took only 3 months. Soon, thanks to the support of the EIC Accelerator, we will be able to tackle a new indication in only a few weeks.
We are developing mCUBE, an Artificial Intelligence (AI)-augmented cloud-based precision oncology platform for multi-omic, multi-cancer and multi-source data identification, aggregation and analysis. mCUBE encapsulates Epigene Labs’ deep expertise in oncology, data science, and bioinformatics into a user-friendly solution that will make a difference for cancer patients. Our solution delivers 3 main gains:
- increased productivity for our users (cancer scientists),
- increased profitability for our customers (biopharma companies), and
- improved care for cancer patients ultimately.
The conventional timeline for discovering a new target for a specific indication is 2 years. At Epigene, in the context of our first partnership with the Eyquem’s lab at UCSF, we not only rediscovered known antigens but also prioritised a list of promising targets for relapsed/refractory multiple myeloma in only 6 months. A few months later, in a second program focusing on colorectal and gastric cancer, this same task took only 3 months. Soon, thanks to the support of the EIC Accelerator, we will be able to tackle a new indication in only a few weeks.