Mots-Clés

spatial transcriptomics visium HD tumor micro-environment LUAD cancer-associated fibroblasts machine learning AI/Deep learning python R bash

Description

Scientific Context

Lung cancer remains the most frequently diagnosed cancer worldwide with 2.5 million new cases reported each year (Bray et al., 2024, DOI: https://doi.org/10.3322/caac.21834). Among its subtypes, lung adenocarcinoma (LUAD) is the most prevalent, accounting for approximately 45% of all cases. Current therapy strategies combine surgery, radiotherapy, chemotherapy and/or targeted or immunotherapy, depending on tumor stage and molecular characteristics. Recent studies have demonstrated that the tumor microenvironment (TME) – mainly composed of fibers, immune cells and cancer-associated fibroblasts (CAFs) – plays a critical role in tumor progression and can influence therapeutic response (Sahai et al., 2020, DOI: https://doi.org/10.1038/s41568-019-0238-1). Through single cell transcriptomics data analysis, our team identified distinct CAF subtypes, two of which are associated with T-cell exclusion (Grout et al., Cancer Discovery 2022, DOI: https://doi.org/10.1158/2159-8290). Among these, the MYH11+ CAF subtype, detected only in early stage LUAD, forms one layer of elongated cells lining tumor cells, as seen by multiplex histochemistry imaging. This spatial organization suggests a direct interaction with tumor cells and the potential role of this CAF subtype in LUAD progression. To further elucidate the molecular and spatial features of this CAF population and its crosstalk with adjacent tumor cells, our team generated spatial transcriptomics (ST) state-of-the art technologies, including sequencing-based (Visium HD) and imaging-based (MERFISH and Xenium) approaches.

Main goals

The goals of this intership project are to: 

• optimize the internal Visium HD ST pipeline for the detection of CAF subtypes:

• Benchmark spatially aware tools for data quality controls, normalization, and clustering

• Compare binning versus pseudo-cell strategies for data processing

• Implement both manual and automated approaches for cell type annotation

• investigate cell-cell communication, particularly between CAFs and tumor cells:

• identify potential ligand-receptor interaction pairs involved in CAF-tumor crosstalk

• identify transcription factors and signaling pathways driving these interactions

Candidate Profile

We are looking for a M2 student with:

• a solid background in Bioinformatics and a strong knowledge in statistics and machine learning (Deep learning is a plus)

• proficiency in Bash, Python and R programing languages (Nextflow is a plus)

• strong interest in Biology and Oncology

• prior experience in single cell or spatial transcriptomics data analysis

• good communication skills and the ability to present scientific approaches withing a multidisciplinary research environment.

Research Environment

The internship will be conducted at the Institut Curie Research Center, in the Stroma and Immunity team led by Dr. Hélène Salmon. Institut Curie offers a unique environment combining a leading cancer Hospital with a top-tier Research Center, facilitating close collaboration between basic research and clinical applications, including access to primary tumor samples. The Salmon Lab is part of the “Immunity and Cancer” Department (Inserm U932), headed by Ana-Maria Lennon-Duménil, including experts in cell biology, immunology, clinical immunotherapy, as well as a large working group of bioinformaticians. The Salmon laboratory combines wet lab and computational approaches to understand the contribution of stromal cells, especially CAFs, on shaping immune responses against cancer. The M2 student will also be part of the Curie Bioinformatics HUB, a dynamic structure supported by the Curie Bioinformatics Core Facility (led by N. Servant, P. Hupé and E. Barillot), that promotes best practices and peer exchange among bioinformaticians. 

Supervision:

The Stroma and Immunity team is composed of 3 postdocs, 1 PhD student, 1 lab manager, and 3 bioinformaticians. The M2 students will be supervized by both H. Salmon and M. Andrianteranagna (senior bioinformatician in the team).