Computational Biology of Cancer

The “Computational Cancer Biology” research team, within the “Cancer Data Science” unit affiliated with IHU-Prism, aims to identify intrinsic and extrinsic tumor determinants of malignant initiation and progression through the analysis of high-dimensional data associated with detailed clinical annotations. Our team focuses on characterizing tumor genomics and intrinsic cellular programs, as well as the role of the tissue microenvironment, in the initiation and progression of the disease. We integrate multiple levels of the tumor phenotype—such as genomics, transcriptomics, and the tissue context—with a particular focus on the interactions between tumor cells and the normal cells in their environment. Using state-of-the-art profiling technologies and machine learning tools, we develop methodological frameworks to study cancer as an evolving ecosystem. Our goal is to unravel the complexity of cancer biology through studies closely linked to clinical practice, in order to develop tools for predicting tumor progression and to identify new therapeutic vulnerabilities.

Our main projects focus on:

• Spatially Resolved Pan-Cancer Atlas of Invasion

  As part of the IHU-Prism initiative, this project aims to decipher the mechanisms—both common to all cancer types and specific to each—that drive malignant invasion within its tissue context. We are conducting large-scale spatial transcriptomic profiling of in situ and invasive lesions, as well as associated stromal regions (immune and fibrotic), in 500 patients, using paired data from H&E staining, DNA sequencing, and proteomics, as well as histological and clinical annotations. This work aims to address a central question in cancer research: What are the driving forces behind malignant invasion—the first step in the malignant cascade—that are common to cancer types with well-characterized precancerous lesions? The objective of this study is to develop new strategies to intercept the disease before it progresses.

• Prediction and Reprogramming of Neoplastic Adaptation

  Metastatic progression requires significant cellular adaptation. This project aims to identify the intrinsic and extrinsic forces that shape the tumor phenotype and determine disease progression, by integrating intrinsic tumor trajectories and the tumor microenvironment. We will trace the phenotypic trajectories linking the primary tumor to metastatic lesions and explore behaviors at the cellular population level that support tumor progression and enable long-term adaptation. To achieve this, we will develop methods that jointly integrate cell lineage tracing, transcriptomic states, functional specialization, and interactions with the microenvironment through multimodal data integration and machine learning models.