Mots-Clés
Single-cell multiome
scRNA-seq
scATAC-seq
Trisomy 21
Gene regulatory networks
eRegulons
Neurodevelopment
Description
CONTEXT
The team studies how genetic variation disrupts gene regulatory networks in neurological and neurodevelopmental disorders (Delahaye-Duriez et al., 2016 Genome Biol ; Dufour et al., 2024 Brain Behav Immun ). Building on open-source datasets, it has assembled a consensus single-cell multiome atlas of the human cortex. The atlas covers 8 public studies, 144 donors from 6 post-conception weeks to 60 years, and more than 1 million nuclei. It is summarised as eRegulons: a transcription factor (TF), its enhancers and its target genes, per cell type and developmental stage. In the CORTEXSCOPE project, the atlas is used to interpret non-coding variants in neurodevelopmental disorders, and the best candidates will be validated by genome editing in iPSC-derived cortical organoids. However, organoid fidelity has mostly been assessed on the transcriptome, not on gene regulation. Before editing an enhancer, one must know whether the organoid reproduces the relevant cell type, stage and eRegulon.
Trisomy 21, the leading genetic cause of intellectual disability, provides a benchmark. It increases the dosage of several chromosome 21 TFs (OLIG2, RUNX1, ERG, ETS2, GABPA, BACH1, SIM2), and should therefore shift identifiable regulatory programmes, both in vivo (Lattke et al., 2026 Nature Medicine) and possibly in organoids (GEO GSE305145).
OBJECTIVES
This internship will establish an eRegulon-level map of what cortical organoids capture of human cortical gene regulation, and use it to assess which trisomy 21-associated alterations observed in vivo can be studied in organoids. The student will:
1. Integrate single-cell multiome data (RNA and chromatin accessibility from the same nucleus) of cortical organoids, from public and/or in-house datasets, with the reference atlas, and map organoid cells onto atlas cell types and developmental stages.
2. Evaluate the fidelity of each eRegulon in organoids (TF expression, accessibility of its enhancers, expression of its target genes) and produce a map of the cell types, stages and regulatory programmes that organoids faithfully reproduce.
3. Characterise trisomy 21-associated regulatory alterations in two in vivo datasets: fetal cortex between 10 and 20 PCW (15 trisomic fetuses, 15 controls) and postnatal prefrontal cortex between 0 and 3 years (5 matched pairs, about 221,000 nuclei), with a focus on eRegulons driven by chromosome 21 TFs.
4. Determine which of these alterations fall within the regulatory space that organoids reproduce, in order to guide the design of future genome-editing and disease-modelling experiments.
The student will acquire strong expertise in integrative genomics and will be part of an ongoing project with direct experimental follow-up. This work will be supervised by Dr Adrien Dufour and Pr Andrée Delahaye-Duriez at the NeuroDiderot laboratory in Paris. The applicant will have the opportunity to gain a high level of competence in the highly demanded areas of single-cell multiomics, gene regulatory network analysis and neurodevelopmental integrative genomics.
PROFILE
M2 student in bioinformatics, computational biology, data science or a related field.
• Interest in biological data analysis and solid background in statistics.
• Proficiency in R or Python in a Linux environment.
• Prior experience in single-cell analysis (Seurat, Signac, Scanpy, ArchR) is a plus.
• Interest in gene regulation and neurodevelopment.
• Knowledge of reproducible-research tools (Git, Conda/Docker, Snakemake, Jupyter notebooks) desired.