Mots-Clés
Machine learning, artificial intelligence, single-cell RNA sequencing, neuron–astrocyte communication, metabolic pathways, cell-cell interactions, neurodevelopmental disorders
Description
Internship title:
Machine learning approaches to infer neuron-astrocyte communication from single-cell transcriptomics
Keywords:
Machine learning, artificial intelligence, single-cell RNA sequencing, neuron–astrocyte communication, metabolic pathways, cell-cell interactions, neurodevelopmental disorders
Internship description:
Recent advances in single-cell transcriptomics provide unprecedented opportunities to investigate how different brain cell types communicate in physiological and pathological conditions. In particular, interactions between neurons and astrocytes are increasingly recognized as essential for brain development, homeostasis, and function, and their disruption may contribute to neurodevelopmental disorders.
Our project brings together expertise in neuroscience, single-cell transcriptomics, bioinformatics, and artificial intelligence. A research engineer within the team has developed computational tools aimed at identifying pairs of neurons and astrocytes that are likely to interact, based on the expression of genes involved in defined metabolic pathways. These approaches provide a first framework to predict neuron-astrocyte functional coupling directly from single-cell transcriptomic data.
The main objective of this M2 internship will be to use, evaluate, and refine these machine learning-based approaches, with the aim of improving the identification of neuron-astrocyte pairs and assessing how predicted interactions are modified following experimental perturbations.
The student will work on single-cell RNA-sequencing datasets generated in mouse models of neurodevelopmental disorders, including control and experimentally manipulated conditions. The internship will focus on developing analytical strategies capable of determining whether metabolic coupling between neurons and astrocytes is altered in response to these perturbations.
More specifically, the student will:
• Apply, refine and optimize existing machine learning and AI-based tools developed within the team to single-cell transcriptomic datasets,
• integrate gene sets and pathway information related to major metabolic processes involved in neuron-astrocyte coupling,
• evaluate the robustness and biological relevance of model predictions,
• compare predicted interaction patterns between control and experimentally manipulated conditions,
• contribute to the improvement, documentation, and generalization of the computational workflow.
Candidate profile:
We are primarily looking for a student with a strong background in machine learning, artificial intelligence, data science, or computational biology.
The candidate should have:
• solid knowledge of ML concepts and commonly used computational approaches,
• good programming skills, preferably in Python and/or R,
• experience with data analysis and high-dimensional datasets,
• the ability to develop, evaluate, and improve computational models,
• an interest in biological applications of machine learning.
Previous experience with single-cell transcriptomics is welcome but not mandatory. Conversely, some knowledge of, or a strong interest in, neuroscience, brain development, or neurodevelopmental disorders will be highly appreciated.
By the end of the internship, the student will have gained experience in applying and refining machine learning approaches for biological data, analysing large-scale single-cell transcriptomic datasets, integrating biological pathway information into computational models, and investigating cell-cell communication in the context of neurodevelopmental disorders.
Some background work can be found in:
Zhao W., Johnston K.G., Ren H., Xu X., Nie Q. “Inferring Neuron-Neuron Communications from Single-Cell Transcriptomics through NeuronChat.” Nature Communications 14 (2023): 1128.
https://doi.org/10.1038/s41467-023-36800-w
Marcy G., Foucault L., Babina E., et al. “Single-Cell Analysis of the Postnatal Dorsal V-SVZ Reveals a Role for Bmpr1a Signaling in Silencing Pallial Germinal Activity.” Science Advances 9 (2023): eabq7553.
https://doi.org/10.1126/sciadv.abq7553
Alghamdi N., Chang W., Dang P., et al. “A Graph Neural Network Model to Estimate Cell-Wise Metabolic Flux Using Single-Cell RNA-Seq Data.” Genome Research 31 (2021): 1867–1884.
https://doi.org/10.1101/gr.271205.120
Host laboratory (name, director and address):
Inria Project-Team AIstroSight
Pierre Wertheimer Neurology Hospital
59 Boulevard Pinel
69500 Bron, France
https://team.inria.fr/aistrosight/
And
Equipe BrainGuard
Centre de Recherche en Neurosciences de Lyon
CH Le Vinatier - Bâtiment 462
Neurocampus, 95 Bd Pinel,
69500 Bron, France
Hosting teams: AIstroSight; BrainGuard
Internship supervisor (HDR):
Dr. Hugues BERRY
hugues.berry@inria.fr
Dr Olivier RAINETEAU
Olivier.raineteau@inserm.fr